When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Well the Kokeln T-bar delete definitely lowered the rear RC.
Because of the semi-trailing arm geometry, lowering the rear by 1" only drops the rear RC by 0.5" whereas at the front lowering the car by 1" drops the front RC by about 1.5-2.0".
That's the basic difference between semi-trailing arm and strut suspension.
Since even at stock height the front RC height is much lower than the rear, lowering the car results in a front RC about 5" lower than the rear. Addressing this requires lowering the rear RC and raising the front RC.
Glad you liked NZ!
Cheers, Mike
Bruce's T-bar delete raises the inner trailing arm attachment points, which should raise the RC:
I don't believe Bruce's bar changes the outer attachment points - someone please correct me if this is wrong. I believe the whole premise behind Kokehln's and Bruce's modification to the rear inner attachment points is as follows: On a lowered car, both front and rear roll centers will drop more than the CG does, creating longer roll couples at both ends. In the front, this can be corrected by using longer ball joint pins, allowing the A-arms of a lowered car to be put back to horizontal, which will significantly raise the front RC. I was under the impression that this would leave the front RC higher than the rear (although I've never seen it quantified as to how much), which is why higher rear inner attachment points are needed to rise the rear RC to match the front (or at least reduce the amount by which the front and rear RC's are out of kilter after the installation of the longer ball joint pins). If someone has a different understanding, please chime in.
So, getting back to my original question of how raising the front and rear RC's on a lowered car impacts the suspension dynamics, Eric K stated earlier in this thread that his car exhibited a greater tendency to understeer after he installed Bruce's bar in his track 968 as a result of the increased grip in the back. Could someone please explain this? Is it simply a result of the fact that the car will tend to sway less because of the shortened lever arm represented by the roll couple, and this increased the size of the contact patches of Eric's rear tires, giving him more grip in the corners? Or is the improved rear grip caused by a different mechanism? Thanks.
My wheel rate calculations are correct. You left off the spring inclination angle correction term, which is the cosine of the angle of inclination of the spring from vertical:
I measured the inclination of my front springs to be 16 degrees from vertical, which only reduces the wheel rate by a factor of 0.96, so it's a small correction, and it's applied equally to the front and rear wheel rates (I didn't measure the rear spring angle, but it looks close to the front's, and it's a small correction anyway), so it doesn't change the conclusions.
Wheel rate is spring rate x motion ratio squared x spring inclination factor. You haven't squared the motion ratio, hence previous poster was correct in saying your wheel rate calculation is wrong.
Bruce's T-bar delete raises the inner trailing arm attachment points, which should raise the RC:
I don't believe Bruce's bar changes the outer attachment points - someone please correct me if this is wrong. I believe the whole premise behind Kokehln's and Bruce's modification to the rear inner attachment points is as follows: On a lowered car, both front and rear roll centers will drop more than the CG does, creating longer roll couples at both ends. In the front, this can be corrected by using longer ball joint pins, allowing the A-arms of a lowered car to be put back to horizontal, which will significantly raise the front RC. I was under the impression that this would leave the front RC higher than the rear (although I've never seen it quantified as to how much), which is why higher rear inner attachment points are needed to rise the rear RC to match the front (or at least reduce the amount by which the front and rear RC's are out of kilter after the installation of the longer ball joint pins). If someone has a different understanding, please chime in.
So, getting back to my original question of how raising the front and rear RC's on a lowered car impacts the suspension dynamics, Eric K stated earlier in this thread that his car exhibited a greater tendency to understeer after he installed Bruce's bar in his track 968 as a result of the increased grip in the back. Could someone please explain this? Is it simply a result of the fact that the car will tend to sway less because of the shortened lever arm represented by the roll couple, and this increased the size of the contact patches of Eric's rear tires, giving him more grip in the corners? Or is the improved rear grip caused by a different mechanism? Thanks.
I wouldn't read too far into a single person's experience of before/after handling changes and draw too many conclusions from that. He might have also changed his rear (or front) alignment settings, spring/damper rates, etc. Even if nothing else changed, some other aspect of his base setup could have dictated the tendency to understeer (spring rates, tire size, heat cycles, etc). All other things remaining equal, correcting the rear roll center on a lowered car will lessen body roll at the rear and generally make it more stable and predictable as the wheel camber and toe changes while cornering will be less.
On my car (nearly 2" lower than stock) with 2" drop pins in front and stock geometry in the rear, the front end is locked in, but the rear feels a half-step behind. I can feel that it rolls more. But, it is progressive and predictable so not a massive concern. But corrected rear suspension geometry would be the icing on the cake . . . .
It sounds like you already have a decent setup and a guy who knows what he is doing. Just get some Tarett adjustable sway bars, which should be able to help fine-tune the handling balance to compensate for any other changes.
Wheel rate is spring rate x motion ratio squared x spring inclination factor. You haven't squared the motion ratio, hence previous poster was correct in saying your wheel rate calculation is wrong.
Not to nitpick, because we're talking about very small differences, but I did square the motion ratio. My front springs are 700#, so using a front MR of 0.92, and a spring inclination of 16 degrees (the cosine of which is 0.96), the front wheel rate comes out to 700*(0.92)^2*0.96 = 569, which is what I had in post. In the rear, it's 800*(0.65)^2*0.96 = 324, which is also what I had in my post. But it's good that we're all checking each others' work!
Last edited by Cloud9...68; 12-18-2017 at 11:10 PM.
I wouldn't read too far into a single person's experience of before/after handling changes and draw too many conclusions from that. He might have also changed his rear (or front) alignment settings, spring/damper rates, etc. Even if nothing else changed, some other aspect of his base setup could have dictated the tendency to understeer (spring rates, tire size, heat cycles, etc). All other things remaining equal, correcting the rear roll center on a lowered car will lessen body roll at the rear and generally make it more stable and predictable as the wheel camber and toe changes while cornering will be less.
On my car (nearly 2" lower than stock) with 2" drop pins in front and stock geometry in the rear, the front end is locked in, but the rear feels a half-step behind. I can feel that it rolls more. But, it is progressive and predictable so not a massive concern. But corrected rear suspension geometry would be the icing on the cake . . . .
It sounds like you already have a decent setup and a guy who knows what he is doing. Just get some Tarett adjustable sway bars, which should be able to help fine-tune the handling balance to compensate for any other changes.
Chris,
How do your car's front and rear ride heights compare to mine, now that I've measured them in the manner prescribed in the workshop manual? I suspect we're fairly close.
Overall, it seems to me that if someone lowers his car, then flattens his A-arms (and tie rod bars) by using longer ball joint pins and tie rod connectors, and simultaneously raises his rear RC by installing Bruce's bar, the net effect, at least in terms of tendency to understeer vs. oversteer, should pretty well be a wash. So, to a first approximation, assuming the car is well behaved before these adjustments to the roll centers, no adjustments to spring rates, sway bar stiffness, compression/rebound setting, etc. should be needed. Then, as you suggest, the system can be fine-tuned with some adjustable sway bars. Anybody see it differently? Thanks.
Andy, like I mentioned over email, I think rc 'correction' opens a can of worms. You cannot make inferences about dynamic balance based on static analysis of the geometry. Changing the pickup locations will alter camber/caster/toe curves and how rc behaves dynamically (ie, planar shift). Modeling all of these factors completely still doesn't suggest how to fix any problems that arise for obtaining predictable handling in the end.
As mentioned by Cloud and way back on page 3 of this thread, when I first installed the bar in 2013 and made no other changes, the car went from neutral to significant understeer. I was trying to only change one thing at a time. That indicated to me rear grip was up. Lap times also went down, which is the only thing that really matters. The car has gone thru a lot of changes since then.
The bar raises both the inside and outside points, just like the kokeln bar did. The latest version with the upgraded pivot bolt, thicker steel and additional attachment points should be as strong as the kokeln bar.
When you drastically lower the rear, you end up with too much negative camber that can’t be easily adjusted out. The bar fixes that. If you update the front and the rear you should be able to lower the car and maintain suspension angles reasonably close to what they were from the factory.
Andy, like I mentioned over email, I think rc 'correction' opens a can of worms. You cannot make inferences about dynamic balance based on static analysis of the geometry. Changing the pickup locations will alter camber/caster/toe curves and how rc behaves dynamically (ie, planar shift). Modeling all of these factors completely still doesn't suggest how to fix any problems that arise for obtaining predictable handling in the end.
This is a good point. I used the term "correction" in my previous post to imply that the both Bruce's and the Kokeln rear end (along with the drop pins I have in the front) return the 944 static suspension geometry more or less to factory parameters. Whether this is 100% ideal or not is open to debate and depends on the ride height and specific setup in question, but it certainly seems to be an improvement over slamming the suspension and doing nothing about the geometry . . . .
At the end of the day, however, a MacPherson strut front end and semi-trailing arm rear end are far from ideal in terms of wheel control on a race car. Dwain from Vision Motorsports (he made all of the Kokeln stuff) was at my shop a few months ago, and I asked him if he had any of the old 944 rear ends laying around, or at least the tooling, and he did not. He said that the best option was to adapt the 996 rear subframe/suspension like he did on Tony G's race car because the multi-link geometry is so much better. He also added that it was "not that difficult" to install . . . . easy for him to say as he owns/runs a race fab shop!
My advice would be to stick to your plan, buy one of Bruce's last parts (and Tarett sway bars!) and install it when you are ready. Keep driving the car and analyzing data, get some good coaching, and figure out what works for you and your car at the tracks that you run. Only make one suspension change at a time if you can help it!
But the important thing is to have fun---it is easy to get caught up in lap times and niggling little handling problems and forget why you took up track driving in the first place!
This is a good point. I used the term "correction" in my previous post to imply that the both Bruce's and the Kokeln rear end (along with the drop pins I have in the front) return the 944 static suspension geometry more or less to factory parameters. Whether this is 100% ideal or not is open to debate and depends on the ride height and specific setup in question, but it certainly seems to be an improvement over slamming the suspension and doing nothing about the geometry . . . .
At the end of the day, however, a MacPherson strut front end and semi-trailing arm rear end are far from ideal in terms of wheel control on a race car. Dwain from Vision Motorsports (he made all of the Kokeln stuff) was at my shop a few months ago, and I asked him if he had any of the old 944 rear ends laying around, or at least the tooling, and he did not. He said that the best option was to adapt the 996 rear subframe/suspension like he did on Tony G's race car because the multi-link geometry is so much better. He also added that it was "not that difficult" to install . . . . easy for him to say as he owns/runs a race fab shop!
My advice would be to stick to your plan, buy one of Bruce's last parts (and Tarett sway bars!) and install it when you are ready. Keep driving the car and analyzing data, get some good coaching, and figure out what works for you and your car at the tracks that you run. Only make one suspension change at a time if you can help it!
But the important thing is to have fun---it is easy to get caught up in lap times and niggling little handling problems and forget why you took up track driving in the first place!
Chris,
Well put. I've shared Trevor's (PorscheG96) concerns over opening a can of worms by moving the semi-trailing arm attachment points, which is why I kicked off the recent discussions in this thread. But from the experiences of seasoned racers like Eric K and West Dillard (and several others, I'm sure), it does appear that the rear roll center correction that Bruce's bar provides appears to be a net positive. Raising the front RC on strut type suspensions is relatively common practice, and Bruce's and Kokehln's approach for the rear does seem to balance out front and rear roll couples, apparently without significant side-effects. And I've said earlier, I do think that monitoring the impact of the "optimized" roll centers, and the responses to incremental tweaks to other suspension parameters, will improve my feel for what the car is communicating to me.
As mentioned by Cloud and way back on page 3 of this thread, when I first installed the bar in 2013 and made no other changes, the car went from neutral to significant understeer. I was trying to only change one thing at a time. That indicated to me rear grip was up. Lap times also went down, which is the only thing that really matters. The car has gone thru a lot of changes since then.
The bar raises both the inside and outside points, just like the kokeln bar did. The latest version with the upgraded pivot bolt, thicker steel and additional attachment points should be as strong as the kokeln bar.
When you drastically lower the rear, you end up with too much negative camber that can’t be easily adjusted out. The bar fixes that. If you update the front and the rear you should be able to lower the car and maintain suspension angles reasonably close to what they were from the factory.
Thanks Eric, I had made some assumptions. Never seen either bar in the flesh. Apologies for the confusion everybody!
Cheers,
Mike
On a lowered car, both front and rear roll centers will drop more than the CG does, creating longer roll couples at both ends.
Sorry to be pedantic, however this isn't true for semi-trailing arm rear suspension. The rear RC drops a percentage of the ride height drop. This percentage is roughly (track)/(2x effective swing arm length).
For a swingaxle car, the RC height change is the same as the ride height change.
Your diagram shows this (great diagram BTW!) - if you imagine that the drop in IC is created by dropping the ride height (instead of by changing the inclination of the arm), then clearly the IC height change (= ride height change) is less that the RC height change.
However as you point out, the bar seems to be a great success, which is what counts after all!
Cheers,
Mike
As mentioned by Cloud and way back on page 3 of this thread, when I first installed the bar in 2013 and made no other changes, the car went from neutral to significant understeer. I was trying to only change one thing at a time. That indicated to me rear grip was up. Lap times also went down, which is the only thing that really matters. The car has gone thru a lot of changes since then.
The bar raises both the inside and outside points, just like the kokeln bar did. The latest version with the upgraded pivot bolt, thicker steel and additional attachment points should be as strong as the kokeln bar.
When you drastically lower the rear, you end up with too much negative camber that can’t be easily adjusted out. The bar fixes that. If you update the front and the rear you should be able to lower the car and maintain suspension angles reasonably close to what they were from the factory.
Eric,
Are you sure Bruce's bar raises both the inner and outer attachment points? I haven't seen one in person, so I'm only going by the pictures, but it isn't clear to me how it raises the outer attachment point. If both attachment points are raised (the same amount), the roll center wouldn't be raised, would it? Wouldn't that defeat the whole purpose? Thanks.
Eric,
Are you sure Bruce's bar raises both the inner and outer attachment points? I haven't seen one in person, so I'm only going by the pictures, but it isn't clear to me how it raises the outer attachment point. If both attachment points are raised (the same amount), the roll center wouldn't be raised, would it? Wouldn't that defeat the whole purpose? Thanks.
Look at the 2nd picture in the 1st post of this thread, it clearly shows the attachment point raised from the center of the torsion bar 'hole' in the mount.
Here is my crudamentary drawing, as I understand it. The red dots represent the stock suspension and the green dots the raised points of Bruce's bar. The green CG represents the lowered center of gravity because of the car being lower. The distance between the CG and RC will be reduced, reducing roll. These aren't massive changes. More like attempts to get the benefits of lowering the car without altering how the original suspension works.
Sorry to be pedantic, however this isn't true for semi-trailing arm rear suspension. The rear RC drops a percentage of the ride height drop. This percentage is roughly (track)/(2x effective swing arm length).
For a swingaxle car, the RC height change is the same as the ride height change.
Your diagram shows this (great diagram BTW!) - if you imagine that the drop in IC is created by dropping the ride height (instead of by changing the inclination of the arm), then clearly the IC height change (= ride height change) is less that the RC height change.
However as you point out, the bar seems to be a great success, which is what counts after all!
Cheers,
Mike
Mike,
Not at all - you've made great contributions to this discussion, which have sent me scrambling to google and youtube, which I appreciate, because the process is helping me understand this stuff better. Just one clarification, though. In the second sentence of your last post, you say that the RC at the back drops a percentage of the ride height drop. By a percentage, I assume you mean the RC drops by an amount that is somewhat less than the ride height drop. Assuming the drop in CG is equal to the ride height drop, this means that unlike at the front, the roll couple is actually shorter at the rear after the car is lowered than it was at its previous higher ride height.
But then near the end of your post, you say that the IC height change, which is equal to the ride height change, is less than the RC height change, which would mean that the RC drops by more than the change in ride height. I assume you meant to say that the IC height change is more than the RC height change, correct?
Look at the 2nd picture in the 1st post of this thread, it clearly shows the attachment point raised from the center of the torsion bar 'hole' in the mount.
Here is my crudamentary drawing, as I understand it. The red dots represent the stock suspension and the green dots the raised points of Bruce's bar. The green CG represents the lowered center of gravity because of the car being lower. The distance between the CG and RC will be reduced, reducing roll. These aren't massive changes. More like attempts to get the benefits of lowering the car without altering how the original suspension works.
This is an excellent clarification - thanks. I was under the mistaken impression that Bruce's bar raises the rear RC by increasing the angle of inclination of the semi-trailing arm by raising the inner attachment point relative to the outer, but I see that I was mistaken - the picture you point out int Bruce's first post clearly shows the outer attachment point being raised. And your diagram clearly shows the means by which Bruce's bar actually raises the RC. It makes sense now.
So, let's see if I have this all straight:
- Lowering the car lowers the front RC more than the front CG, increasing the front roll couple, potentially by quite a bit in a significantly lowered car
- The same lowering actually lowers the rear RC less than the rear CG, creating a slightly shorter roll couple in the rear
- The front roll couple can easily be corrected by moving the control arms back to horizontal, which significantly shortens the front roll couple, probably to where it's shorter than it was as stock height
- Bruce's bar further shortens the rear roll couple (but not by a drastic amount), bring the front and rear roll couples into reasonable balance
12-18-2017, 01:45 AM
