ARB LLT

[salad]

Hi guys, first of all, I have searched, and couldn't find what I was after.

Now, I've currently been working on spreadsheets for our 08 car and I'm a little bit stuck on the LLTs for ARBs at the moment.

In Milliken, they've just treated ARBs as extra roll stiffness and have treated the elastic LLT the same way as if the total roll stiffness was entirely made up from springs.

I've always been under the belief that ARBs work by transferring load from the inside wheel onto the outside wheel, therefore they should have a different effect on LLT compared to springs.
This can be seen in practice with a car running big ARBs, they always pick up the inside wheel.

What I've currently done is use the stiffness of the ARB, the amount of twist on the bar, and calculate the forces coming off the lever arms and then put this into my LLT calculations.
This seems to make sense to me, but other members of my team are saying it's wrong.

Just like some people's opinions on here.

Cheers
Showza

[ben]

The Milliken treatment of ARBs is correct.

The lateral load transfer is a function of lateral acceleration, CG height and track width. The springs and bars have nothing to do with the total load transfer.

What the springs and bars do affect is the distribution of load transfer front to rear. If you make one end of the car stiffer than the other it will take more of the lateral load transfer. The big ARB you're talking about will just be transferring a large enough proportion of the load to unload the inside wheel, you could do the same with springs alone.

Ben

[salad]

Sorta got the idea, PM sent about a couple of little things...

[Marshall Grice]

The only thing I would add to Ben's description is that the sway bars are also able to transfer unsprung weight where as springs alone can not (well not in a typical coil spring configuration where the spring is unattached in full droop)

[salad]

Transfer unsprung weight? I think this might be what is causing my confusion.

Right now I'm wondering what happens to the forces that come off the ARB lever arms. These forces would go through the bellcranks and then into the push/pull rod, and then would affect the wheel load, right?

Sorry if my questions seem repetitive, I'm just really struggling to get my head around it at the moment.

Cheers
Showza

[Marshall Grice]

typically coil springs are shorter then the full droop length of the shock, in other words when the car is in the air the spring has been off the seat for a while.
once the spring has reached its free length the force on the spring is zero. the only weight on the tire is the unsprung weight. the only way to actually lift the tire off the ground with out a swaybar is for the shock to extend to full droop. typically this full droop of the shock is well past the normal operating range of the suspension.
of course it's possible to have the spring not go free before the shock is fully extended, it's just not normal. it's also possible to use different type of springs (torsion bars) that wouldn't go free in full droop.

the sway bar links are 2 force members so there isn't the 'zero spring rate' area near the end of the travel. which is why you will normally see cars with sway bars lifting wheels more often then cars without, but as Ben said you can do the exact same thing with springs alone.

other then that a sway bar is nothing but a spring and the analysis is just as in RCVD. You can convert the sway bar torsional rate into a vertical wheel rate and calculate the roll stiffness accordingly.

[jdstuff]

Originally posted by Marshall Grice:
of course it's possible to have the spring not go free before the shock is fully extended, it's just not normal.

I'm sorry, but I whole-heartedly disagree with this statement. In all of the formula cars that I have worked on (or around), more often than not the spring does not go free at full droop. Often the springs were set up to either just "touch" at droop, or with a preloaded spring. Sometimes they would be preloaded so much that it would cause a zero-droop suspension setup. <gasp!>

Granted, there were applications that warranted running rattle in the springs....I'm not saying that never happens. But even then sometime a light-rate (think 4-20 lb/in) "helper" spring would be added in series to primary coil spring. So even then, if the car hit full droop there would still be a spring force applied to the tire.....albeit a small one.

Saying that the "normal" for a formula car is to have a free spring a full droop is just flat out incorrect. There is no normal setup, and thinking so is quite narrow minded.

Sorry for the rant....

[Marshall Grice]

so shoot me for using the wrong word. http://fsae.com/groupee_common/emoticons/icon_smile.gif

I would agree that it is much less common to see a shock with a loose spring at full droop on a formula car then other cars but it still happens.

I mean come on, did anyone else design their suspension to have the shock in the middle of it's stroke only to realize that the spring compressed much less then half the stroke with weight on wheels after it was built? http://fsae.com/groupee_common/emoticons/icon_redface.gif
wait i mean uhmmm...change spring rates to a stiffer then designed spring and have to lower the perch height to maintain proper motion ratios. yeah that's it.

[murpia]

Originally posted by Marshall Grice:
I mean come on, did anyone else design their suspension to have the shock in the middle of it's stroke only to realize that the spring compressed much less then half the stroke with weight on wheels after it was built? http://fsae.com/groupee_common/emoticons/icon_redface.gif
wait i mean uhmmm...change spring rates to a stiffer then designed spring and have to lower the perch height to maintain proper motion ratios. yeah that's it.
Generally, if you have a longer damper stroke in droop than required for your 'usable wheel travel' and you need to 'clock' the rockers to get the motion ratios correct, you should use either helper springs or droop limiters.

An exception might be a 3rd spring arrangement on an aero car. Then, you need to design the spring seats properly so any spring gap is contained correctly and doesn't allow the spring to 'catch' anywhere it shouldn't.

Regards, Ian

[salad]

Thanks for the input guys, but it seems that the main question I'm interested in still has not been addressed (or quite possibly I'm an idiot and havn't understood the explainations properly).

Quoting from Milliken (p281);
"Twisting of the bar adds load to one wheel and removes it equally from the other."
Would it be safe to assume that this "load transfer" would be proportional to the stiffness of the ARB?

Now I'm quite content to accept that total LLT is fixed, and a bigger ARB on one end changes the distribution of this, BUT what if the ARBs' stiffness' were increased equally on both sides such that the LLTD remains the same, what happens to all this extra "load transfer" from the stiffer ARBs?

Once again, sorry if this has been explained and I simply havn't understood. If someone could spell it out, that would be great http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cheers
Showza

[ben]

Originally posted by salad:
Quoting from Milliken (p281);
"Twisting of the bar adds load to one wheel and removes it equally from the other."
Would it be safe to assume that this "load transfer" would be proportional to the stiffness of the ARB?

Now I'm quite content to accept that total LLT is fixed, and a bigger ARB on one end changes the distribution of this, BUT what if the ARBs' stiffness' were increased equally on both sides such that the LLTD remains the same, what happens to all this extra "load transfer" from the stiffer ARBs?


If you stiffen both bars by the same amount the LLTD stays the same but the roll gradient decrease - i.e. less roll angle for a given lateral acceleration.

What would be different with the stiffer bar is that if you hit a bump with one wheel you will transfer load off the other wheel more rapidly but this is a transient and not relevant to calculating roll stiffness distribution for steady-state cornering.

Ben

[flavorPacket]

salad, in the case you described, there is no 'extra' load transfer from stiffer ARBs. Load transfer is a function of wheelbase, cg height, and force. You cannot alter the total load transfer with ARBs, only the way it is distributed.

[salad]

Originally posted by ben:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by salad:
Quoting from Milliken (p281);
"Twisting of the bar adds load to one wheel and removes it equally from the other."
Would it be safe to assume that this "load transfer" would be proportional to the stiffness of the ARB?

Now I'm quite content to accept that total LLT is fixed, and a bigger ARB on one end changes the distribution of this, BUT what if the ARBs' stiffness' were increased equally on both sides such that the LLTD remains the same, what happens to all this extra "load transfer" from the stiffer ARBs?


If you stiffen both bars by the same amount the LLTD stays the same but the roll gradient decrease - i.e. less roll angle for a given lateral acceleration.

What would be different with the stiffer bar is that if you hit a bump with one wheel you will transfer load off the other wheel more rapidly but this is a transient and not relevant to calculating roll stiffness distribution for steady-state cornering.

Ben </div></BLOCKQUOTE>

That makes sense for single wheel bump, but the quote that I took from Milliken was for roll. The preceding sentence is "No twist of the torsion bar takes place if the wheels move up and down together (ride), but in roll the bar is twisted as one wheel moves down and the other up from some initial position."

So, what is the effect of this load that they are talking about in roll?
What am I missing that makes this not affect wheel loads?

Greatly appreciating everyone's input so far http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cheers
Showza

[Marshall Grice]

I think you have your thinking all backwards. The swaybar doesn't 'generate' any load, it reacts to the load transfer generated from the lateral acceleration.