Ok, Old Europe'ers. STOP GIVING WRONG REAR SWAY BAR INFO
Moderators: The Dark Side of Will, Series8217
-
- Peer Mediator
- Posts: 15629
- Joined: Wed Nov 24, 2004 11:13 pm
- Location: In the darkness, where fear and knowing are one
- Contact:
Yes, BUUUUT...
A car with soft springs and big bars is an story of compromise.
Dampers must be matched to springs. Heavy roll bars add roll stiffness with no corresponding increase in roll damping.
Such cars have little body roll in the corners, but much up and down suspension motion over heaves and dips and bottom out easily.
Corvettes are frequently pointed out as having this short coming when compared to true no-holds-barred sports cars like Ferarris.
Bars should be a method of fine-tuning the suspension, not the primary means of providing roll stiffness. As such they should not get terrible large.
Certain cars (like the Morgan Aero do not use bars at all, yet still handle quite well.
A car with soft springs and big bars is an story of compromise.
Dampers must be matched to springs. Heavy roll bars add roll stiffness with no corresponding increase in roll damping.
Such cars have little body roll in the corners, but much up and down suspension motion over heaves and dips and bottom out easily.
Corvettes are frequently pointed out as having this short coming when compared to true no-holds-barred sports cars like Ferarris.
Bars should be a method of fine-tuning the suspension, not the primary means of providing roll stiffness. As such they should not get terrible large.
Certain cars (like the Morgan Aero do not use bars at all, yet still handle quite well.
- Series8217
- 1988 Fiero Track Car
- Posts: 5989
- Joined: Thu Jun 02, 2005 9:47 pm
- Location: Los Angeles, CA
yup. with the sway bar, the ups & downs gets loaded onto both springs. when you add a rear sway bar to Fiero, the rear gets really stiff. but, thats probably why we like it so much. covers up the crappy springs & struts.Series8217 wrote:So when we increase bar size/stiffness, we need to appropriately change damping as if we increased the spring rate?
We don't care about body roll and if a camber or toe change is so offensive that we don't want near it, we install larger bump stops to prevent the car from entering it.
You're over-thinking this. It's really simple.
The amount of lateral load transfer that a car has is based on three things: height of the center of gravity, G and track. The equation is (height X G) / Track This is a fraction of total weight.
The amount of weight transfer longitudinally is (height X G) / length
A fiero's wheelbase is roughly 90 inches. Its track is roughly 60. I've rounded the numbers to simplify the math.
So it looks like that.
If we guess the height of the fiero's CG to be 30 inches off of the ground and say we can generate 1 Gs on sticky summer rubber we get:
1/3 vs 1/2. We need the anti-roll bars to make up extra spring-rate while cornering for the 1/6th of extra weight transfer we are generating laterally to prevent bottoming of the suspension.
This is assuming we're not doing dumb stuff like moving the roll centre up and generating jacking forces.
You're over-thinking this. It's really simple.
In the 60s, Mark Donahue hypothesized that there was a thing called the friction circle. The theory was that a tire could generate approximately the same G braking as cornering and if using 75% of available traction for braking, there would be 25% left over for cornering and vice versa. Today we can quantify this to the thousandth of a pound in traction force for specific tires in specific directions and combinations, but the theory still holds roughly true.Series8217 wrote:Ooo Oo me me!!
To reduce body roll..
for the purpose of... preventing severe camber changes?
The amount of lateral load transfer that a car has is based on three things: height of the center of gravity, G and track. The equation is (height X G) / Track This is a fraction of total weight.
The amount of weight transfer longitudinally is (height X G) / length
A fiero's wheelbase is roughly 90 inches. Its track is roughly 60. I've rounded the numbers to simplify the math.
So it looks like that.
If we guess the height of the fiero's CG to be 30 inches off of the ground and say we can generate 1 Gs on sticky summer rubber we get:
1/3 vs 1/2. We need the anti-roll bars to make up extra spring-rate while cornering for the 1/6th of extra weight transfer we are generating laterally to prevent bottoming of the suspension.
This is assuming we're not doing dumb stuff like moving the roll centre up and generating jacking forces.
- Shaun41178(2)
- Posts: 8464
- Joined: Fri Nov 19, 2004 7:12 pm
- Location: Ben Phelps is an alleged scammer
-
- Peer Mediator
- Posts: 15629
- Joined: Wed Nov 24, 2004 11:13 pm
- Location: In the darkness, where fear and knowing are one
- Contact:
Really depends on what kind of engine you've got inside. 30 might be high but it doesn't change the point.
The Dark Side of Will wrote:eHoward wrote:If we guess the height of the fiero's CG to be 30 inches off of the ground and say we can generate 1 Gs on sticky summer rubber we get:
I've heard 19 inches...
- Series8217
- 1988 Fiero Track Car
- Posts: 5989
- Joined: Thu Jun 02, 2005 9:47 pm
- Location: Los Angeles, CA
So if the front suspension was right at its limit of travel while braking at 1G.. then without swaybars we would be bottoming the suspension (in front at least) if cornering at 1G due to the increased lateral weight transfer compared to the longitudinal weight transfer experienced while braking?eHoward wrote:...
1/3 vs 1/2. We need the anti-roll bars to make up extra spring-rate while cornering for the 1/6th of extra weight transfer we are generating laterally to prevent bottoming of the suspension.
...
So if, without swaybars, we are not bottoming our suspension while cornering at the maximum, then adding them would be pointless? Or do we still want to make 1G braking cause the same weight transfer front to back as 1G cornering does side to side for other reasons?
I'm assuming the actual weight being transfered (the lbs, not the percentage) would be different for each end of the car without 50/50 weight distribution.
Since a Fiero has more weight in the back, the weight being transferred longitudinally would be different than the weight being transferred transversely for the same G force being applied either way.
Since a Fiero has more weight in the back, the weight being transferred longitudinally would be different than the weight being transferred transversely for the same G force being applied either way.
Hold on here. That was to explain at the most basic level why we run anti-roll bars from side to side and not from front to back.
I'm going to add to this post. The thing is, and this is where Donahue's theory becomes valueble, we aren't just pure braking or pure cornering. Most times, we're doing both.
I'm going to add to this post. The thing is, and this is where Donahue's theory becomes valueble, we aren't just pure braking or pure cornering. Most times, we're doing both.
Series8217 wrote:So if the front suspension was right at its limit of travel while braking at 1G.. then without swaybars we would be bottoming the suspension (in front at least) if cornering at 1G due to the increased lateral weight transfer compared to the longitudinal weight transfer experienced while braking?eHoward wrote:...
1/3 vs 1/2. We need the anti-roll bars to make up extra spring-rate while cornering for the 1/6th of extra weight transfer we are generating laterally to prevent bottoming of the suspension.
...
So if, without swaybars, we are not bottoming our suspension while cornering at the maximum, then adding them would be pointless? Or do we still want to make 1G braking cause the same weight transfer front to back as 1G cornering does side to side for other reasons?
- Series8217
- 1988 Fiero Track Car
- Posts: 5989
- Joined: Thu Jun 02, 2005 9:47 pm
- Location: Los Angeles, CA
Mark Donohue! I have respected that name for over 30 years. I read "The Unfair Advantage" as a kid, and now have a hardback copy of it.
The friction circle in 2 dimensions would translate, I am thinking, into a force vector in a particular direction at any one time. The total combination of acceleration or braking in the +/- Y direction say, and the lateral cornering in the +/- X direction. So in effect, the tire traction at any one time would be limited to a total vector in some direction from the contact patch.
Is that the theory??
So now, I am thinking, there is also a +/-Z vector, which is the force of gravity under static conditions. Braking, accelerating, and cornering all alter the Z vector on any one tire at a given moment.
Are there computer simulation programs where all three force vectors could be taken into account? I am thinking one could start with the cars basic suspension component data (Weight, springs, anti-way bars, control arm moment arms, etc., etc.), and then set up a set of dynamic conditions (speed, curve radius, steering angle, etc.).
Then possibly one could predict at what point tire break-away occurs to some degree (maybe +/- 5%). Just a thought.
The friction circle in 2 dimensions would translate, I am thinking, into a force vector in a particular direction at any one time. The total combination of acceleration or braking in the +/- Y direction say, and the lateral cornering in the +/- X direction. So in effect, the tire traction at any one time would be limited to a total vector in some direction from the contact patch.
Is that the theory??
So now, I am thinking, there is also a +/-Z vector, which is the force of gravity under static conditions. Braking, accelerating, and cornering all alter the Z vector on any one tire at a given moment.
Are there computer simulation programs where all three force vectors could be taken into account? I am thinking one could start with the cars basic suspension component data (Weight, springs, anti-way bars, control arm moment arms, etc., etc.), and then set up a set of dynamic conditions (speed, curve radius, steering angle, etc.).
Then possibly one could predict at what point tire break-away occurs to some degree (maybe +/- 5%). Just a thought.
To reply to your first question, tire behavior is funny, but that's the general idea.
Weight transfer and downforce are what alter "Z". In college, I used Adams Car to model it. http://www.mscsoftware.com/products/pro ... cfm?PI=424
this is actual tire data showing "Z"
Weight transfer and downforce are what alter "Z". In college, I used Adams Car to model it. http://www.mscsoftware.com/products/pro ... cfm?PI=424
this is actual tire data showing "Z"
A. Frayn wrote:Mark Donohue! I have respected that name for over 30 years. I read "The Unfair Advantage" as a kid, and now have a hardback copy of it.
The friction circle in 2 dimensions would translate, I am thinking, into a force vector in a particular direction at any one time. The total combination of acceleration or braking in the +/- Y direction say, and the lateral cornering in the +/- X direction. So in effect, the tire traction at any one time would be limited to a total vector in some direction from the contact patch.
Is that the theory??
So now, I am thinking, there is also a +/-Z vector, which is the force of gravity under static conditions. Braking, accelerating, and cornering all alter the Z vector on any one tire at a given moment.
Are there computer simulation programs where all three force vectors could be taken into account? I am thinking one could start with the cars basic suspension component data (Weight, springs, anti-way bars, control arm moment arms, etc., etc.), and then set up a set of dynamic conditions (speed, curve radius, steering angle, etc.).
Then possibly one could predict at what point tire break-away occurs to some degree (maybe +/- 5%). Just a thought.
eHoward: Thanks for the software link. It looks like a complete set of modules on that package would cost more than a car buildup.
As suspected, non-linear behaviour as a function of Z versus traction. I guess one would assume a constant road to tire friction coefficient.
I may try to contact some old college buddies who stayed there to get advanced degrees and teach. Sometimes the university is willing to grant donating alumni a few liberties.
stimpy: Its a little engineering-speak. I remember just enough to translate the concepts. I had 4 years of vector analysis, and then some, rammed down my gullet studying flight dynamics.
As suspected, non-linear behaviour as a function of Z versus traction. I guess one would assume a constant road to tire friction coefficient.
I may try to contact some old college buddies who stayed there to get advanced degrees and teach. Sometimes the university is willing to grant donating alumni a few liberties.
stimpy: Its a little engineering-speak. I remember just enough to translate the concepts. I had 4 years of vector analysis, and then some, rammed down my gullet studying flight dynamics.