Sandro

iTrader: (1)

This discussion is getting more interesting by the day.

Max, I believe the assumptions in your example are incomplete when you say that "the car still rolls 4 degrees, since we didn't change anything that would affect the total amount of body roll"

Since you have stiffened the front and softened the rear, you will now actually have less wheel travel in the front and more wheel travel in the rear. The 300 lb weight transfer still stays. You will now have 300/175=1.7" in the front and 300/125=2.4" in the rear.

Sandro

iTrader: (1)

I am ready to re-write history immediately...

I rushed writing my previous post but I was wrong. Max is right. Since the chassis does not flex, you must have the same 2" travel F and R.

Therefore, sway bars (as well as springs) do influence the weight transfer split between F and R (and not just the speed as I previously stated).

- Sandro

2MCHPWR

iTrader: (2)

whats it called when you disconnect your front sway bar entirely to get better rear traction via more/quicker weight transfer when drag racing?

sandro, why did you write the chassis does not flex? i think mine does a lot i want a rollbar.

Sandro

iTrader: (1)

I was just referring to Max' example, when he said that both F and R move by the same 2". Of course the chassis does flex in reality. But this should not be taken into account in Max' elegant explanation on how the weight gets transfered to the end with the higher roll stiffness.

As of why you do disconnect the F sway bar when drag racing, I don't know. It shouldn't matter since you go straight and most of the the weight is on the rear as you keep accelerating. But maybe it is because you don't always go perfectly straight and some weight does move to the front when you shift. By disconnecting the front sway bar you would reduce roll stiffness in the F and minimize the chances that relatively more weight get transfered to a side front wheel. I guess by the same token you should also use very soft spring in the F to keep the roll stiffness to a minimum. But this is just a wild guess of mine. I have zero experience in drag racing and as I may have noticed I am quite confused in general...

- Sandro

DamonB

Correct. The swaybar cannot change the total amount of weight transfer across the car, but it can change how that weight is split front vs. rear and therefore change the fundamental front/rear balance of the car. This is what all of us with adjustable or non-stock bars are doing.

Your copies must be getting dusty. Crack 'em open every once in a while.

DamonB

The engine is bolted to the chassis. During launch the chassis must resist the torque of the engine and in doing so the chassis rolls since it rests on top of springs. This roll is resisted by the springs and bars.

With the front swaybar in place the chassis is forced to stay flatter as the car launches and this makes it harder to get the front wheels into the air since there is more roll resistance that must be overcome before a front wheel goes airborne. Why do we want the front wheel(s) in the air at launch? The car weighs a given amount and there is a given amount of load transfer to the rear at launch. If one or both of the front wheels are in the air that means more load is resting on the rear tires and they can offer the maximum amount of grip possible. Even if the front wheels don't actually leave the ground removing the front bar still allows the front end to roll more freely and therefore transfer more weight across the rear vs the front. More weight transfer to the rear during launch means more grip at the rear.

If the front tires are in the air 100% of the load must be on the rear tires. The car doesn't become lighter with the fronts airborne, so all the weight must now be at the rear! If the springs and bars could change the total amount of weight transfer as some would have you believe then any time you lifted a tire from the ground your car would become lighter. This is of course impossible. All the springs and bars can do is adjust the front vs rear split, but the total is always the same.

Lowering the front spring rate would accomplish the same goal as disconnecting or removing the front sway bar.

Sandro

iTrader: (1)

You are absolutely right. As I said, it was many years ago... Evidently, I must have lost lots of neurons in the meantime... Fortunately, it looks like I am still able to properly rewire the few left, although with noticeable time and effort...



By the way, your post #9 - why do you say (referring to speed of weight transfer)
"No they are not; shocks do that. Springs and swaybars have absolutely no effect on how quickly the weight transfers across the car."

I don't think you are correct here, all suspension components do, in particular springs, as it take time for them to compress to the point where all load is transferred to the tires - although it wouldn't surprise me to be wrong again...

- Sandro

maxcooper

I cheated and called my brother, who is both a racer and a dynamics engineer at a major automaker, for some supporting info. He helped come up with the example that I posted. I don't have the book, but off the top of his head, he suggested that I look at Milliken (RCVD) p683.

-Max

Sandro

iTrader: (1)

Too modest. Your example was simple to grasp and powerful in its conclusions.

By the way, I have been moving or adding roll stiffness in the front and got interesting results. By moving 150 lb/in wheel rate to the F [zero left in the rear], the entire 600 lb weight transfer would add 600 lb load to the poor outside F tire [I guess this would be true in a three wheel car, or in a four wheels car with the chassis connected to the rear suspension/wheels trough a single pivoting point intersecting the roll axis]. You would get a similar result with a super stiff front (massive anti sway bar or super stiff springs), with negligeable roll this time.

Interestingly, with my stock autox car (stock springs but 1.25" solid front sway bar) I have been using 265 mm tires in the F (and spacers) and 225 mm in the R, and still experiencing hotter temperature in the F tires. I guess I better understand now why.

- Sandro

KevinK2

Such a 4 wheel car exists. The last triumph spitfires and GT6's had a swing axle rear suspension, with rear differential mounted to frame, and a pivoting connection between top of diff and the transverse leaf spring. This was to eliminate rear wt transfer due to roll, and stop the dangerous jacking effect.

Note this thread has been about wt transfer due to roll only. There is also transfer due to unspung wt, and more importantly transfer due to the front or rear sprung weight as a shear load at the front or rear roll center.

For the triumph, little rear wt transfer due to roll, but high rear roll center (top of diff) caused significant transfer due to rear sprung weight. My old gt6 had race springs up front, 1" front bar, and a custom rear adjustable bar I designed (no oem rear bar) and welded in. My analysis set rear bar at only 9/16" with long arms (easily twisted by hand), but it was perfect to tweak fr/rr weight transfer .... neutral position for track events, 1 step stiff for autox.

A solid 32mm front bar (racing beat, suspension techniques), in roll, has an effective rate at the springs of 1600 lbs/in. Monster bar.

https://www.rx7club.com/forum/showth...light=sway+bar

DamonB

During a transition the springs compressing or extending is proof that the weight transfer has already happened. The load is already present, the springs compressing or extending is a result of that. It's not chicken or egg; the load transfer happens and the springs extend or compress in proportion to the change in load.

Springs react instantly to the magnitude of the load; they have no care for the speed at which that load is applied. A shock on the other hand is velocity sensitive and reacts in direct response to how quickly the load occurs, not necessarily to the magnitude of it.

KevinK2

I see it differently.

Cut the steering wheel on a mush suspension, and the car cg does not instantly follow the wheel path ... sprung weight shear reactions at roll centers do near instantly load outside wheels. But loading due to spring compression takes a finite amount of time .... you can not have the high vertical load on the outside tires unless the link to the sprung weight ... springs ... are compressed. The inertia due to body roll would likely spike the weight transfer before steady state loads occur.

With stiff suspension, cg does follow wheel path from the start, more rapid weight transfer due to slight body roll, less of a spike in outer tire contact loads due to body roll inertia.

jkstill

iTrader: (11)

From GRM article cited earlier in the thread:
(http://grmotorsports.com/news/012005...l-bars.php]GRM article)

"Because a firmer anti-roll bar allows less deflection, it will transfer side-to-side energy (lateral loads) at a faster rate."

jgrewe

iTrader: (1)

See any post above that refers to shocks/dampers.

I can have a shock that is valved so tight that it does more to the car entering a turn than the spring. Oh yea, thats what DamonB(and I) have been getting at.

DamonB

KevinK2, you're confusing weight transfer with response. A stiffer spring setup will produce faster response than a softer spring, but the speed of weight transfer among the tires will be the same in both cases.

The pause in the car CG following your control input is due to the chassis heaving around and taking a moment to set to a steady state. This does not effect how fast the weight transfer occurs, it effects the response time of the chassis. Two different things.

If the car has shock absorbers on it it can.

The shocks determine how quickly the spring is allowed to react to the force input. This means you can get load increase at the tire without compressing the spring. Sit a shock/spring assembly on a scale and strike it with a hammer. The scale will register the large load increases of each hammer blow even though the spring compresses barely a fraction.

Shocks work by the fact that they force the springs' response to be slowed and out of phase compared to the input. The loads can reach the tire through the shock without the spring displacing and the magnitude of that load at any instant will be dependent on the shock valving.

As long as the suspension is moving and shocks are in use tire loading is not in direct proportion to spring displacement.

KevinK2

DamonB,

My experience with stock and some Koni shocks, various cars, is that compression damping is much lower than rebound. Yes they will contribute to the transient loading and unloading of tires. True race shocks can have a huge effect.

But, with modest/oem shocks, spring stiffness does effect wt transfer rate. Extreme case of rigid springs, no shocks needed, turn in and bang, wt transfer done. Repeat with old dodge van (no sway bars) ... watch the clock as the van body rolls.

Hammer test on strut ***'y can be misleading, as there is a significant preload to overcome.

"This means you can get load increase at the tire without compressing the spring" I could buy this on extension of stock shocks, not compression.

I agree that shocks influence initial wt transfer. My initial post was isolating spring stiffness only. Considering shocks, it is another variable. Most oem shocks will not mask wt tranfer response change if roll rates are increased 4X, as springs and bars can do. But, with stock springs and the stiffest shocks, shocks will determine wt transfer rate.

Sandro

iTrader: (1)

Kevin,

Thank you for your post #35 and the reference in it. Great information! Anyhow, I have been perfectly happy with my 32mm bar in autox even though I realize only now how stiff the front roll is with a bar like that.

As for the weight/load transfer through the springs, I believe we are on the same page. It takes more time for the tire to be loaded through softer springs than stiffer ones. Softer springs need to be compressed more to develop the same force required to balance the weight transfer, and this takes more time. With a solid suspensions or equivalent (e.g. a super stiff sway bar) load is transfered instantaneously (I am glad I seem to be right this time...)

- Sandro

jgrewe

iTrader: (1)

You might be confusing how long it takes for the car to lean over with soft springs(it leans more) with weight transfer.

Sandro

iTrader: (1)

Maybe the terminology I have been using is not right.

What I am trying to say is that it takes time for the tire to get fully loaded. I don't beleive the tire is loaded istantaneously with soft suspenstions. At least for the portion concerning the load transfered through body roll. I believe the load on the tire increases as the spring compresses. Since the car rolls more with softer springs (less roll resistance), it takes more time for the spring to get fully compressed. Since the load transmitted by the spring is proportional to the amount of compression, the tire gets loaded more as the spring keeps compressing.

- Sandro

DamonB

Yes. Confusing "weight transfer" with "chassis response".

This is true of any shock (that isn't ****); they all have more rebound. In compression the shock and spring are working together against the load but in rebound the shock must do all the work otherwise the spring is left to itself and oscillates, so rebound forces must be much stronger.

It's not misleading at all. Unscrew the hat and remove all the preload. Your results will be the same. Heck, take the spring off and hammer or push on the shock rod. Same result. The fact that shocks exert a force against the suspension is exactly what a shock dyno plot shows us and shocks are always dynoed without springs attached.

It's true in both compression and rebound, though most of the rate is added in rebound. Shock dyno of some shock:



The diagram above is a shock dyno plot from two different shocks. The vertical scale is pounds of force (pounds of resistance the shock exerts) and the horizontal scale is shaft velocity. There is no spring on the shock for this test. The Koni Sport in this test is a rebound only adjustable shock. The shaded blue area illustrates the rebound adjustment range between minimum and maximum. You can see that at maximum rebound a shaft velocity of 2 in/sec produces 275 pounds of rate. For a shaft velcoity of 2 in/sec in compression the rate is only about 20 pounds.

Only 20 pounds of added rate in compression isn't much, but 275 pounds of rebound is a lot! Rebound is what really makes the chassis work and is what drivers really feel (Koni yellows off the shelf are rebound only adjustable for instance). It's easy to focus on the load on the outside tires in the corner but inside tires are just as important! The car will have more overall grip if we can keep the tires more evenly loaded. Weight transfer decreases total grip, that's why all racecars are as low as possible.

At a shaft speed of 2 in/sec the outside tire in our example only gains 25 additional pounds of spring rate in compression, but the inside tire gains 275 pounds(!) of additional rate in rebound that resists the chassis rolling in the first place because the inside suspension is not allowed to go into droop as quickly as it would like. If the inner suspension can't go into droop the chassis can't roll. This keeps more load on the inside tire. The inside suspension is just as important as the outside in resisting roll! The more rebound we run the longer it will take the weight to transfer off the inside tire because the rebound prevents the inside suspension (and thus the tire) from unloading as quickly as it would like. This keeps weight on the inside tires for a longer period of time while the chassis is rolling.

The shocks do not change how much weight transfer there is, but they can be used to control how long it takes for that weight transfer to happen. The other thing to keep in mind is shocks only generate forces when the shaft is moving. Anytime shaft velocity equals zero the shock is doing absolutely nothing. Once the car is keeled over and in steady state again the shocks aren't doing a thing; the car is merely riding on the springs and bars. Once the chassis begins to un-roll the shocks come back into play.

DamonB

You guys are still seeing this backwards. The load on the tires increases and the spring compresses as a result! The spring compressing doesn't do anything to the tire, the spring is merely compressing in response to the load that is ALREADY being transmitted to the tire! As I've attempted to show in my explanation above once the suspension is moving and a shock absorber is in use the amount of load on the tire has more to do with the shocks than the springs anyway.

KevinK2

"This means you can get load increase at the tire without compressing the spring"
"Anytime shaft velocity equals zero the shock is doing absolutely nothing"

Damon, some basic conflict in your statements.

------

The oem curves above are perfect to illustrate my point. At 2"/sec shock (and spring) speeds, you have about 10 lbs in compression, and 20 lbs in extension.

given a .6 motion ratio up front, 600 at the tire static, force at spring is about 1000 lbs. In a corner with 500 lb springs and 32mm front bar, net spring rate in roll is about 2000 lb/in. For say .9" compression in the corner, spring and bar compress .5", with an increase in force of 1000. This is basically full wt transfer with tire going fromm 600 to 1200 lbs. The spring/bar force change from 1000 to 2000 lbs makes the shock resistance at 10 lbs insignificant. Same goes for the extension side vs the 20 lb damping force (note that zero force at tire is due to force equilibrium reached as spring extension is balanced by the sway bar force on arm).

Same conclusion is reached for the Koni's at softest setting, per above curves.

Please note that koni's at full hard is not a linear extrapolation of their excellent rebound response curve ... it is a semi lock-up condition not recommened for use, and forces pulled off that limit curve are misleading.

Here are some better curves:
http://www.fd3s.net/suspension.html#SHO

To elaborate on the extension side, with 250 lb at 2 in/sec for an adjusted Koni. Using the above example, net spring/bar force on the a-arm drops from 1000 lbs to zip. Total droop is .5" at spring and shock.

First .1" of droop .... spring/bar sys energy loss is .5k(.5^2-.4^2) = 90 inlb, while damper absorbes 25 in-lb, so response is mostly due to spring/bar rates initially.

Over total droop, spring/bar releases 250 in-lbs, and shock absorbes 125 in-lb of energy. So shock definitely slows down droop at the end of cycle. In reality, the 2"/sec initial velocity would slow down, and shock would take a little less.

Bottom line, stifferer springs/bars will still make wt transfer quicker, as you have to wait for the outboard springs to compress and most shocks will not influence this. Stiff rebound shocks inboard will temporarily shift roll center inboard, complicating things.

To the extreme, ultra stiff rebound would have full wt transfer done when outer springs are compressed, and cg of car would be lowered due to roll center at inboard tire contact patches. Once in curve, inboard springs would slowly extend to normal cornering height, and cg would also come back up. Weight transfer would not change, ignoring slight cg rise.

Sandro

iTrader: (1)

I don’t think so.

Put a coil spring on the floor and step on it. Your weight is not transferred instantaneously to the floor. The only force the spring can transmit to the floor is proportional to how much it compresses. It takes time for your weight to be fully transferred to the floor. The very same time it takes for the spring to get fully compressed and generate a force equal to your weight. Until then, a lesser force is generated by the spring and this is the only one that can be transmitted to the floor. If the spring is stiffer, it needs to compress less to generate an equal force, which takes less time. Therefore, stiffer springs transfer weight/load faster than soft ones.

An extreme example is this one: toss the steel coil spring, step on a chair and jump on the floor. The only “spring” left between you and the floor is a column of air under your feet. This spring is so soft that no weight/load is transferred to the floor until you land. Still, the air molecules under your sole are subject to your full weight.

I wish Max Cooper or his brother could come up with a better killer example and resolve this discussion for good…Max, are you there?

Back to the steel spring on floor. If you are a Newton’s first law of motion lover and are bothered by the notion that two unequal forces can be applied to the two spring ends, you shouldn’t be. Unbalanced forces are what cause bodies to move, including the top of the spring and the spring’s CG as a whole. The top moves until there is difference between your weight and the force generated by the spring as it compresses. The bottom does not move, therefore the forces must be balanced there all the time. But, since the force exerted by the spring on the floor is also proportional to the amount of spring compression, the force exerted by the floor upon the bottom of the spring can only be an equal but opposite normal force. You could also look at the spring as a whole and consider the forces as applied to the spring GC. The CG gets pushed down by your weight and pushed up by the normal force exerted by the floor. The resulting force applied to the GC is the difference of the two. Your weight exerts a constant force downward; the normal force upward is proportional to the amount of spring compression. Initially, the force applied to the CG is equal to your weight and the GC starts moving down. As the spring compresses and the force generated by it builds up, the resulting force gets smaller and eventually becomes zero. At that point, the normal force is equal to your weight and the spring CG stops moving.

- Sandro

DamonB

A spring has absolutely no idea how fast something is happening and therefore has no effect on the speed of weight transfer. A shock on the other hand IS a velocity sensitive device and so the response from the shock is in direct proportion to how quickly the input was applied.

I've been giving poor explanations I guess. I'll look for some others' words who know how this all works. Milliken & Milliken I'm sure has excellent definitions of such things as "response" but I don't have mine in front of me at the moment.

The links below as well as Milliken and Millike and Brian Beckman's Physics of Racing do excellent jobs of describing these things. Again we can go all the way back to the equations that calculate the percentage of weight transfer during cornering:

Lateral Load transfer= (G*CG height)/track width

Where
G = cornering G
CG = center of gravity

If you're not changing cornering G, CG height, or track width you're not having any effect at all on how much weight your car transfers across the chassis! There is no hidden meaning in any of this. Springs, bars, shocks or tires do not change the total amount of weight transfer! End of story.

http://www.afcoracing.com/tech_pages/shocks.shtml

"Rebound control is a shock's resistance to extend. The amount of rebound control developed by a shock will affect how quickly the tire is unloaded during dynamic weight transfer and how quickly the suspension "rebounds" or returns to its original position, after the spring has been compressed."

"The rate at which a tire is loaded or unloaded during dynamic weight transfer is affected by the low piston speed control of the associated shock. In rebound, a stiff shock slows down and a soft shock speeds up the unloading process."

This example in regards to circle track (left turn):
"The balance of traction between the left side and right side tires determines to a great extent how the car will handle while decelerating through the corner. For example, a race car will tend to push (not turn) whenever the left side tires do not have enough influence in stopping the car (the right side tires are slowing the vehicle more than the left so the vehicle tends to go to the right). By using stiffer shocks (especially a stiffer extension (rebound) control on the left rear, and to a lesser degree, a stiffer extension (rebound) control on the left front), the unloading process of the inside tires (due to dynamic weight transfer) to the outside tires slows. Consequently, the left side tires remain loaded further into the corner which helps to turn the chassis."

http://www.koni.com/_cars/_general_i...ndersteer.html

"Note that the shock absorbers do not change the amount of weight transfer, only the time it takes to transfer this weight."

http://www.koniracing.com/ovalsetup.html

Again in reference to a left turn:
"The left side rebound settings should be used to control weight transfer to the right side of the car. Shocks do not change the amount of weight transfer, only the time it takes to transfer the weight."

http://www.smithees-racetech.com.au/...shocktune.html

"The stiffer shock always transfers weight faster than the softer shock, in both bump and rebound."

"The stiffer shocked wheel pair will always transfer more weight than the softer shocked wheel pair (just as for wheel pair stiffness from springs, bars and suspension geometry). Each shock adds to wheel pair stiffness, whether in bump or rebound. Shock forces add to roll and pitch resistance."

"In rebound, it is very important to visualise how the rebound shock force is adding to wheel pair stiffness as well. The spring is letting the weight go, but the shock is being extended by the chassis movement. The force it generates is resisting chassis roll and pitch. "

"A stiffer shock transfers the same amount of weight as a softer shock."

"In the last phase of roll or pitch, the stiff shock in rebound is still moving, while a softer shock would have already stopped. The chassis is still moving. Won't the stiff shock still be transferring weight? No, it will not. The weight has already been fully transfered. The wheel loading is constant.

" The Most Common Mistake: "Less front rebound allows for a greater amount of weight transfer to the rear under acceleration." We know this is wrong. The same amount of weight will transfer for stiff or soft rebound in the front."

http://www.smithees-racetech.com.au/...hocktune1.html

"The damper velocities and travel directions resulting from each
cornering phase affect the distribution of load among the four tires."

http://www.turnfast.com/tech_handlin..._springs.shtml

"Note: many people are under the misconception that body roll causes weight transfer. This is not true. See the weight transfer article for details about this."

http://www.turnfast.com/tech_handlin...eightxfr.shtml

"Contrary to what you may be inclined to believe, the amount of weight transfer is not altered by springs, shocks, anti-roll bars, etc."

"The rate of weight transfer impacts the responsiveness of the car to driver inputs....It turns out that shocks have the largest impact on rate of weight transfer."

DamonB

I've listed a lot of links above and put as much effort into this as I'm willing to. All of those links are worth reading. You can bring a horse to water but you can't make him drink.