[Turninconcepts.com]

Originally Posted by Susp Tuner

Hey Turninconcept

First of all thank you guys for your warm welcome. I appreciate it.

How soft you go is a function of three things. How bumpy the circuit is, how the suspension geometry reacts and what the tyre wants.

I've always been of the view you run as soft as you can. This helps the car ride the bumps and keeps the contact patch load as uniform as possible. If I'm dealing with a circuit that is pretty smooth i'll be thinking low speed damping ratios of 0.7 and the high speed and rebound in the order of 0.4. However if I know I'll be running on a goat track I'll tend toward the number on my paper because I know that road holding starts to take precedence, and If I need to work the tyres hard I'll do it with the bars.

That being said the Subaru WRX is an intresting case. Being a road car it has a high centre of gravity and it runs McPherson Struts front and rear. While McPherson struts are OK for a road car, once you start putting race grip into it, the McPherson strut runs out of steam very quickly. This is due to excessive roll centre variation and excessive camber gain. Match that to a tyre that has a hard compound and it needs to be worked hard, then the car needs some order of stiffness to keep this under control. You combine that with a tyre that needs to be worked then you need to go hard in the springing and damping.

How much will be a function of the suspension geometry movement and how the tyres need to be loaded. Typically If you see +/- 20mm that is OK but honestly I won't be nailed down on this figure. The best way to determine it, and I'm not giving anyone a sales plug is to use dynamic race car simulation to determine what the cambers and roll centres are doing and then you can go from there.

As I said before, this is a hard business with no hard and fast rules. I'm not saying that to dodge a bullet, but I am saying that because there are many compromises that must be entailed.

Best Regards

Danny

Actually, that reply is pretty spot on to something Stretch and I were talking about just the other day on the phone.

The McP setup is a bit of a double edged sword I think everyone has found - works great for parts replacement and ease of working on, but as you've pointed out we learned the downsides from a suspension standpoint a while ago.

I think your typical tire you're going to see for performance street driving out here is a decently stiff sidewall for a street tire, with your typical performance all season tire in the winter.

The one thing that we all need to keep in mind (myself included) is that 99% of the time these cars are going to be on the street so we have to take that into account. I'd say that overall the streets tend to be decently smooth with only a few super rough spots (I can think of a few in Chicago. Man, I wish those Chicago guys would just finish with that stupid construction.)

[BigSky]

has anyone ever measured the cg on the Impreza? I know Stretch's calculator has a figure in there, but I think it's a estimate- guess I'm curious to see how the estimate was arrived at (or if in fact was measured)

I've seen some methods of determining cg w/ the use of scales- anyone tried that?

[Turninconcepts.com]

Originally Posted by BigSky

has anyone ever measured the cg on the Impreza? I know Stretch's calculator has a figure in there, but I think it's a estimate- guess I'm curious to see how the estimate was arrived at (or if in fact was measured)

I've seen some methods of determining cg w/ the use of scales- anyone tried that?

Got a link to how to determine that method? I've been kicking around the idea of buying a set of our own scales so I can stop borrowing some...

[BigSky]

here's one Longacre Racing Online -- Tech Article "Center of Gravity Height"

<-- would also like my own set of scales

[stretch]

Originally Posted by Susp Tuner

I've always been of the view you run as soft as you can. This helps the car ride the bumps and keeps the contact patch load as uniform as possible. If I'm dealing with a circuit that is pretty smooth i'll be thinking low speed damping ratios of 0.7 and the high speed and rebound in the order of 0.4. However if I know I'll be running on a goat track I'll tend toward the number on my paper because I know that road holding starts to take precedence, and If I need to work the tyres hard I'll do it with the bars.

What are your general guidelines for what you consider "low speed damping". 50mm/second? 100mm/second? I know it has to be a smooth taper, but I ask because I've seen a bunch of dampers that have their peak damping factor at 130+ mm per second. I don't understand this. They taper off both before and after that point. Can you think of a reason why this would be desirable, or is this just an undesirable byproduct of their bleeder valve?

Here is an example of that, although far from the most extreme one I can find: the coilovers that Turn-In Concepts is developing. The rear dampers (red line) follow your formula, the fronts do not. However, the fronts (blue line) are more typical of what we've seen from the stock dampers and some Ohlins and Bilstein designs.


Any comments on that?

However you do need some tyres that need a bit more of a workout. This could be due to a number of reasons, hardness of the compound, sidewall stiffness. Basically when the tyre needs to be worked that's when you see spring rates rise, and the low speed damping ratios get quite large (in the order of 0.7 to 0.8 sometimes 1.2! that's if the circuit can handle it) A classic example is that when V8 Supercars where running Bridgestone tyres their spring rates where double what they are now.

I'm curious, how does tire rate factor into your damper calculations? If we use the tire rate in series with the main spring, then the combined spring stiffness is less than that of the main spring. Thus, if a damper were valved to have a damping ratio of 0.7 on the main spring, it might be 0.8 when taking the tire into consideration too. I know the tire is self-damped to a small extent, but it doesn't seem like much since a tire bounces if I drop it.

On another message board, someone said not taking tires into account can throw your damping estimates off by 40%. I wish I knew how to verify this, but I'm not really sure what to expect from tires and I don't have the tools to do my own measurements.

Based on measured deflection at rest, I've estimated my tires to be 1400lb/in, although I suspect that is very sharply progressive. The sidewalls are subjectively extremely stiff for a street tire, whatever they measure at, but they're still a street tire.

Any guidance on this subject?

Thanks again for posting here, it's rare to have the opportunity to learn about this stuff.

Originally Posted by Turninconcepts.com

Got a link to how to determine that method? I've been kicking around the idea of buying a set of our own scales so I can stop borrowing some...

If you can angle the car on scales and measure the weight transfer, that'll tell you the CG. You'd be measuring the weight transfer due to CG migration.

My guess was 20 inches, IIRC, and then in my spreadsheet I figured out what the unsprung CG must be to get an overall CG of 20 inches (required for my calcs). It's an estimate of an estimate!

[MGizzle]

Taking tires into the calculations is a pain. Your tire spring rate is different at 30psi and 50psi. So is your tire damping. So as Danny stated, you have to estimate and assume as is the case with many things in tuning a car. I remeber looking at tire data and when some friends tried to use it to account it for dampening. They had to pick a value since it was quiet different for different scenarios. This is where my preaching of POTENTIOMETERS comes in play. You get those suckers on your car and you know what your ENTIRE unsprung mass is doing, including your tire and all the flex in the components. In my opinion all modern race car engineers don't DESIGNfor perfection. They design for something "tunable" or "versatile". You create cars with adjustable pick up points, dampers, springs, etc. etc. so that once you have a good base you use DATA ACQUISITION to get to perfection. This is why I don't like the whole math and calculating part of discussion about chassis tuning. I have seen so many people make claims about their numbers using shady computer simulations and calculations and when they hit the track the data looks like they drove a whole different vehicle.

Anyhow, low speed damping would be qualified as the speeds that the car sees when you roll and pitch etc (no bumps and so on). I was told during my days that 2in/sec and below is low speed. Claude Rouelle used to claim it and I witnessed more than 90%+ percent of our car at that time to be below 2 in/sec, which made sense in autocrossing. If you had your STi equiped with sensors and data acquisition you could see what the damper velocities would be when you are doing a slalom (constant speed etc.). That is what I have used when thinking of tuning for low speed dampening. The high speed stuff is when you hit bumps etc., hence why you damper curve should hit a different slope after a certain input velocity. The Dodge Neon SRT-4 dampers illustrate this drastically. I think I have their damper curve somewhere, I will try to post it.

Anyhow, the Neon curve is higher damped up to some velocity and then the shims in the damper can't move anymore once deflected, giving it lower damping after this happens since more fluid flows by now, which was desing intended. As far as reaching high dampening at 6-8 in/sec, that blows my mind too.

[MGizzle]

oh, as far as determining the CG you can not do it with weight scales. It is BS estimating. You have to lift the front or rear of the car by almost 6-10 feet in order to determine the CG on weight scales in the z-axis.

[stretch]

Originally Posted by MGizzle

Taking tires into the calculations is a pain. Your tire spring rate is different at 30psi and 50psi. So is your tire damping. So as Danny stated, you have to estimate and assume as is the case with many things in tuning a car. I remeber looking at tire data and when some friends tried to use it to account it for dampening. They had to pick a value since it was quiet different for different scenarios. This is where my preaching of POTENTIOMETERS comes in play. You get those suckers on your car and you know what your ENTIRE unsprung mass is doing, including your tire and all the flex in the components. In my opinion all modern race car engineers don't DESIGNfor perfection. They design for something "tunable" or "versatile". You create cars with adjustable pick up points, dampers, springs, etc. etc. so that once you have a good base you use DATA ACQUISITION to get to perfection. This is why I don't like the whole math and calculating part of discussion about chassis tuning. I have seen so many people make claims about their numbers using shady computer simulations and calculations and when they hit the track the data looks like they drove a whole different vehicle.

Do you have any data you can post?

Is it even possible to aim for perfection? The scenario I envision is this: if my tire has a really high rate, it'll finish rebounding after a bump before the rest of the unsprung mass has really had a chance to rebound. This would mean the movement of unsprung mass would not follow a predictable path and would instead have lots of smaller stalls and accelerations- kind of like harmonics in audio. Even if I could use a potentiometer to confirm this, I'm not sure what I'd do about it.

Heck, let's start simple: do you want a higher or a lower damping ratio as you move to a firmer tire? (Or, do you focus more on changing spring rates and keep the damping ratios similar?)

Like I said, I'd love to see any data you have so that I could learn from it. I don't care if it's applicable to my own car, I just think it'd be interesting to see.

Anyhow, low speed damping would be qualified as the speeds that the car sees when you roll and pitch etc (no bumps and so on). I was told during my days that 2in/sec and below is low speed. Claude Rouelle used to claim it and I witnessed more than 90%+ percent of our car at that time to be below 2 in/sec, which made sense in autocrossing. If you had your STi equiped with sensors and data acquisition you could see what the damper velocities would be when you are doing a slalom (constant speed etc.). That is what I have used when thinking of tuning for low speed dampening. The high speed stuff is when you hit bumps etc., hence why you damper curve should hit a different slope after a certain input velocity.

I usually hear below 3in/second but have heard below 5in/second. I would imagine this speed varies with the speed of the steering rack (speed of driver input) as much as it does roll stiffness. The only reason I was wondering is because of the way I see so many dampers valved- with the peak at a much higher velocity than I'd expect to see. I'm happy to hear I'm not the only one puzzled by the 6in/sec or higher peaks, and as a sanity check just wanted to hear it from someone else that 6in/second was never "low speed" for a tarmac car.

[MGizzle]

I just asked for some data from my former Formula SAE buddies. What happened is that a lot of FSAE teams payed Milliken's company some money to do some tire testing for the 10'' and 13'' hoosier, good year tires etc. (slicks). What the test was is pretty much what the OEM's due to test tires. The data is enormous and it takes months or even years to fully grip. I am hoping he can give us some of the data that showed the spring rate etc. response with tire pressure changing.

As far as the low speed damping I will place a call with another friend to confirm but I understand why you are so puzled with the 5in/sec+ damping. Just think about it. If your damper moves 5 inches in one second and MR is about 1 (let us assume this) that means your tire is off the ground in 1 second!!! That is huuuuuuuuuuuge and it can only be caused by a bump (pothole, debries etc.). If you are rooling and you see 5 in/sec the inertia of the car would be enormous since the angular acceleration around the x axis would be tremendous. Man, imagine experiencing those kind of velocities for sports cars is just un acceptable.

I understand that production vehicles are tested for off road terrain and potholes etc. since those can cause some serious loss of control if you are driving at higher speeds but that would lead me that the OEM's are making the damping lower, not higher for these events.

Again, I just think that a lot of places are not sure what they need to be doing with dampers. Another comment is, how is a damper supplier supposed to give you the "perfect" damper when there is thousands of people world wide doing different things to their cars. You might be runing different tires and wheels than I am, or you might be runing different power on your car and the weight might be different. All these factors change the car behaviour quiet a bit and a damper supplier like Ohlins can not account for all that. That is why the damper suppliers make (supposedly) adjustable dampers so everybody can find the sweet spot. I just think it is impossible to get right unless you have EVERYTHING set in your car and you are not changing a thing.

Also, keep in mind we are just talking rolling and bump/reboud scenarios, which are the simple case. What about when you are cornering and steering and braking/accelerating at the same time(some people call this "warp")? Good luck figuring out the damping for that without million dollar budgets like the race car teams.

The perfect set up does not exist. There are so many variables that play into the performance of a car that they are constantly changing (telemetry anyone??). Track temp, ambient temp, car in front of you, car behind you, tire wear, fuel consumption etc. etc. are just SOME variables that change all your equations again...and again...and again...

That is why racing cars is so cool and I love it. Talk about a challenge.

[stretch]

Originally Posted by MGizzle

That is why racing cars is so cool and I love it. Talk about a challenge.

Cheers.

[Turninconcepts.com]

Originally Posted by MGizzle

I just asked for some data from my former Formula SAE buddies. What happened is that a lot of FSAE teams payed Milliken's company some money to do some tire testing for the 10'' and 13'' hoosier, good year tires etc. (slicks). What the test was is pretty much what the OEM's due to test tires. The data is enormous and it takes months or even years to fully grip. I am hoping he can give us some of the data that showed the spring rate etc. response with tire pressure changing.

As far as the low speed damping I will place a call with another friend to confirm but I understand why you are so puzled with the 5in/sec+ damping. Just think about it. If your damper moves 5 inches in one second and MR is about 1 (let us assume this) that means your tire is off the ground in 1 second!!! That is huuuuuuuuuuuge and it can only be caused by a bump (pothole, debries etc.). If you are rooling and you see 5 in/sec the inertia of the car would be enormous since the angular acceleration around the x axis would be tremendous. Man, imagine experiencing those kind of velocities for sports cars is just un acceptable.

I understand that production vehicles are tested for off road terrain and potholes etc. since those can cause some serious loss of control if you are driving at higher speeds but that would lead me that the OEM's are making the damping lower, not higher for these events.

Again, I just think that a lot of places are not sure what they need to be doing with dampers. Another comment is, how is a damper supplier supposed to give you the "perfect" damper when there is thousands of people world wide doing different things to their cars. You might be runing different tires and wheels than I am, or you might be runing different power on your car and the weight might be different. All these factors change the car behaviour quiet a bit and a damper supplier like Ohlins can not account for all that. That is why the damper suppliers make (supposedly) adjustable dampers so everybody can find the sweet spot. I just think it is impossible to get right unless you have EVERYTHING set in your car and you are not changing a thing.

Also, keep in mind we are just talking rolling and bump/reboud scenarios, which are the simple case. What about when you are cornering and steering and braking/accelerating at the same time(some people call this "warp")? Good luck figuring out the damping for that without million dollar budgets like the race car teams.

The perfect set up does not exist. There are so many variables that play into the performance of a car that they are constantly changing (telemetry anyone??). Track temp, ambient temp, car in front of you, car behind you, tire wear, fuel consumption etc. etc. are just SOME variables that change all your equations again...and again...and again...

That is why racing cars is so cool and I love it. Talk about a challenge.

I wholeheartedly, and I think Stretch and 4Banger will also, agree that there is no PERFECT setup.

Although, I think that if you work through it, look at the common variable, choose and use the set up as intended you can get pretty close.

Could a perfect system be done? Sorta. For a given day, course, temp, tire, driving style, etc. In the end though there's always something that can be found to change.

What is nice about going through this entire exercise is that it educates people looking for a good setup (or at least what they feel is a good setup).

[DuncanG]

Originally Posted by MGizzle

As far as the low speed damping I will place a call with another friend to confirm but I understand why you are so puzled with the 5in/sec+ damping. Just think about it. If your damper moves 5 inches in one second and MR is about 1 (let us assume this) that means your tire is off the ground in 1 second!!! That is huuuuuuuuuuuge and it can only be caused by a bump (pothole, debries etc.). If you are rooling and you see 5 in/sec the inertia of the car would be enormous since the angular acceleration around the x axis would be tremendous. Man, imagine experiencing those kind of velocities for sports cars is just un acceptable.

confused and speeling dont help

5in/sec on a road car is nothing. On a cross-country hack I've seen (estimated by video monitoring) 40"/sec compressions and thats with 'overdamped' Ohlins. But of course thats bump damping not roll damping.

Imagine going through a chicane and transferring from 2" droop/bump to 2" bump/droop, so 4" travel. How long do you want to take for that to take a set? Thats mainly a function of the sprung frequency (unless very overdamped) so lets assume its sprung for 2Hz (in heave) and the bars roughly double roll-stiffness so we're talking 4Hz in roll. From trough to peak (or full-left to full-right) is 1/2 cycle so that gives us 0.5 secs for the transition - seems reasonable. Now 4" in 0.5 sec is 8"/sec.

I'm sure Stretch will check whether a 2Hz spring setup is consistent with 2" bump/droop at full tilt but I think its in the ball-park.

[stretch]

Originally Posted by DuncanG

I'm sure Stretch will check whether a 2Hz spring setup is consistent with 2" bump/droop at full tilt but I think its in the ball-park.

I've modeled it as closer to one inch per side depending hugely of course on what sway bars you're running and cornering forces. My model is optimistic since I'm not accurately accounting for bushing flex, but I still think that once your stiff, a significant portion of roll comes from tire deflection instead of the springs.

Doubling the spring rate doesn't double the natural frequency, though. In your example you'd go from 2hz to 2.85hz, roughly. The damped frequency is what counts, though, and it will be lower- probably around 2.3hz. 1/2 oscillation at 2.3hz is a very short amount of time, roughly .22 of a second, so I suppose the potential for high damper speeds exists if you could get the car to shift its weight instantly. There's the catch: I don't think that's possible.

Your speed of transition is going to depend on how fast you can turn the steering wheel and then for the car to actually begin changing direction. For smoother (but still quick) steering inputs, we'll see much lower damper speeds for obvious reasons. Then again, these slower steering inputs don't need much damping anyway.

[4banger]

Take this as a grain of salt, since I did read it on the internet after all. I've seen one or two sources that say 1 in/sec is the cut off for slow speed.

I don't have the answer on where the cut off really is, but I think Optimum G points us in the right direction by talking about transmissibility. The crossover point would probably be a good place to start thinking about where to divide high and low speed damping. This is going to change depending on your spring rates, damping rates, etc, which is probably why there are so many recommendations on where to set the low/high speed switch.

[DuncanG]

Originally Posted by stretch

I've modeled it as closer to one inch per side depending hugely of course on what sway bars you're running and cornering forces. My model is optimistic since I'm not accurately accounting for bushing flex, but I still think that once your stiff, a significant portion of roll comes from tire deflection instead of the springs.

Doubling the spring rate doesn't double the natural frequency, though. In your example you'd go from 2hz to 2.85hz, roughly. The damped frequency is what counts, though, and it will be lower- probably around 2.3hz. 1/2 oscillation at 2.3hz is a very short amount of time, roughly .22 of a second, so I suppose the potential for high damper speeds exists if you could get the car to shift its weight instantly. There's the catch: I don't think that's possible.

Your speed of transition is going to depend on how fast you can turn the steering wheel. For smoother (but still quick) steering inputs, we'll see much lower damper speeds for obvious reasons. Then again, these slower steering inputs don't need much damping anyway.

Back to school for me then