Interconnected suspensions

[747heavy]

.....since a heave spring is not anchored to the chassis like a corner spring.

that depends on your heave spring system, in the photo below, the heave spring would be anchored to the chassis, similar to a monoshock.
But for the function of the heave spring, it does not matter if the spring is anchored at the chassis or just between the two rockers.

[ringo]

@ Ringo
Maybe go and have a look at an F3 (or any other) monoshock suspension.
Here you have no corner springs, and only one heave spring, it may helps to
understand the system of a heave spring without corner springs a bit better/more easily.
Your heave spring, can have any stiffness you want, and can be soft, hard, progressive, linear, digressive - you decide, what you would like it to be.





other then that Scarbs has run an article about "corner springless rear suspension" in F1, maybe you find something there.

http://scarbsf1.wordpress.com/2010/12/0 ... evolution/

I read the article and there are some discrepancies. I am not a suspension expert ,i repeat. But the heave spring is mentioned as acting as the main spring, then it is said that it has a range of free movement.
It cannot have a range of free movement if it is to be used without corner springs.
Th car will roll freely, even with an ARB, until the bumps are hit by the moving geometry providing resistance to the heave spring, "kicking in" or engaging the spring.
I get the intention of removing the corner springs, but it may make more sense if they were simply super soft.
He is right about the drawbacks though.

[ringo]

.....since a heave spring is not anchored to the chassis like a corner spring.

that depends on your heave spring system, in the photo below, the heave spring would be anchored to the chassis, similar to a monoshock.
But for the function of the heave spring, it does not matter if the spring is anchored at the chassis or just between the two rockers.

Ahh good example.
When i say anchored, i mean the resistance is provided by the chassis and a corner of the car. This spring seems to be anchored, but it isn't.
One corner cannot put any force on it, once corner will only cause a rotation and put torsion on the T bar. Both corners, though not equally, are needed to provide a linear motion.
That heave spring is only restrained by both sides pulling on that T bar. And we know the corner springs are providing that pull. Following me?

Now remove the corner springs and imagine what happens in roll.
I think the ARB will be wholly responsible to provide suspension in roll, and it wont have any corner spring to push back against or heave spring.
Only the weight of the unsprung parts.

You see what I am getting at? I am totally open to correction if i am looking at it wrongly.
I'm not being a hardass but i like things laid out clearly before i accept them.

[747heavy]

The only restraint for this spring however (when both wheels are of the ground) is the ARB.

Sorry Ringo, but I don´t think, that this statement is correct.
You can disconnect the ARB and the car will not crash to the ground.
If we talk about a heave spring in between the two rockers.
The distance between the two rocker points will get less in heave, so the spring is restrained in between this two points.

[ringo]

The only restraint for this spring however (when both wheels are of the ground) is the ARB.

Sorry Ringo, but I don´t think, that this statement is correct.
You can disconnect the ARB and the car will not crash to the ground.

I am talking about roll not heave. You will have to have a good imagination to see what i am talking about. Until there is a diagram or someone with suspension software you will get a better picture.
The roll bar doesn't hold up the suspension, i know the car wont crash to the ground without it.

But when the car is in roll, the roll bar will be tied directly to the heave spring, especially with one wheel of the ground.

If we talk about a heave spring in between the two rockers.
The distance between the two rocker points will get less in heave, so the spring is restrained in between this two points.

Ah yes in heave. I think you are following but lagging a bit, now think of roll with no corner springs and a heave spring tying both ends.

Remember my posts are in the context of no corner springs and a constantly engaged heave spring with no free play.

[747heavy]

Yes I think, I see where you getting at.
It comes down to the definition of the modes.
We have pure roll, pure heave and a combination of both.
single wheel bump, is a combination, and the force will come half from the ARB and half from the heave spring.

[747heavy]

yes, in this case, pure roll mode, the length of your heave spring is constant, and there is no force coming from the heave spring (thats why it is called heave spring).
Your force comes from the ARB, what is the problem with that?
Your tire is going to "see" a force, how it is generated is not important for the tire.

If you would run your heave spring inbetween your two rockers, with no corner springs, it would contribute zero force.
If you have no ARB, you would have a car with no stiffness in (perfect) roll.

Still not sure, where we are getting at, with this?
What is the problem, with such a system, and why do you think it would not work?

BTW, I agree that it is highly unlikely, coming back to Scarbs article, that someone would run a setup without "corner springs" and freeplay at the heave spring.
But as I said before, freeplay at the heave spring is not a pre-requisite for such a system. As you see in the photo above, this system has no "freeplay" at the heave spring, just a non-linear springrate.

we had the discussion here on the forum, while talking about ARB´s in F1 cars.
Here a similare layout, as shown above, during single wheel bump.
The heave spring would compress 1/2 the amount, as it would in pure heave mode, and the ARB twists 1/2 the amount it would twist in pure roll mode.
Both "springs (roll&heave)" will contribute force in this case/mode

[Mystery Steve]

I don't want to sway the conversation too far from the intended thread, but something in the picture 747 posted caught my eye. Is that a CF tube used for the ARB? Seems like that would be an odd material choice, if it is what I think it is.

[747heavy]

@Steve

maybe it´s odd, but not impossible.
I know of one DTM/ITC car (1994/95) which run a carbon composite front ARB (for better or worse).

look at the fat black tube near the front bulkhead:



if you don´t mind reading, and are interested in the finer details, maybe this is interesting for you.

http://web.deu.edu.tr/ansys/tezler/yuksek/y12.pdf
http://www.freepatentsonline.com/6431531.pdf

some other applications:

RATH RACING CARBON FIBER TORSION BAR – $240
After three years of racing development in both TT and dirt track, Rath Racing’s Carbon Fiber Torsion Bars are now available to everyone.
These bars come in 22-, 24- and 30-inch lengths for most sport quads, Polaris RZR and Yamaha Rhino applications.
Rath claims these torsion bars react far better than traditional metal bars, and they weigh a whopping 75 percent less.
Used by top pro racers like Harold Goodman and Frank Batista, this lightweight torsion bar teamed with Rath’s carbon fiber sway bar tube is a lethal combination.

there was a time, when F1 teams/engineers where obsessed with CoG height and low weight.
Now, these things may not be that much "en vouge" anymore, and mounting a 1 kg inerter on top of the gearbox is the flavour of the month.
Some years ago, this would have been a no/no, and 20g more per damper the end of the world.
But times and objectives change in racing.

for the carbon fibre fetishists:

MTB coil spring made out of CF.

[DaveW]

Ringo: I think that 747 has posted excellent replies to your questions. If it will help, here is an alternative argument.

Think about a single axle of a fully fitted suspension with unity motion ratios, & imagine that the chassis is fixed in space. Then the vertical loads acting on the wheels are can be calculated from the following relationships (I think):

F1 = (Kh + Kc + Kb)*Z1 +      (Kh - Kb)*Z2
F2 =      (Kh - Kb)*Z1 + (Kh + Kc + Kb)*Z2

where F1 & F2 are individual wheel loads
      Z1 & Z2 are corresponding wheel displacements
      Kh is the stiffness of the "heave", or "3rd" spring
      Kc is the stiffness of each "corner" spring
and   Kb is the stiffness of the anti-roll bar.

Some simple "sanity" checks:

What happens if the displacement of one wheel (Z1, say) is increased (compressed)?

1. With corner springs only: F1 will increase by Kc*Z1, whilst the connected wheel will be undisturbed (OK).
2. With corner springs & a bar: F1 will increase by (Kc + Kb)*Z1, whilst F2 will reduce by Kb*Z1 - i.e. the bar will act to try to lift the connected wheel (OK).
3. With corner springs & a "heave" spring: F1 will increase by (Kh + Kc)*Z1, whilst F2 will increase by (Kh)*Z1 - i.e. the heave spring will act to push down on the connected wheel (OK).

What happens with all springs fitted?

F1 will increase by (Kh + Kc + Kb)*Z1, whilst F2 will change by (Kh - Kb)*Z1 - i.e. the connected wheel will tend to be lifted if Kh is less than Kb, but will be pushed down if Kh is greater than Kb. If Kh = Kb, then the function will be similar to the no bar case (the connected wheel will be undisturbed).

Finally, what happens if no "corner" springs are fitted?

F1 will increase by (Kh + Kb)*Z1, whilst F2 will change by (Kh - Kb)*Z1 - similar in principle to the all springs case, since Kc does not affect the load acting on the connected wheel (for my simple test set-up).

Putting the argument in another way, the addition of a bar will increase the overall "axle" roll stiffness from a base value provided by the corner springs. Removing the corner springs will reduce the base value to zero, allowing the overall "axle" roll stiffness to be set at any value - provided only that there is no limit to the strength or displacement of the bar (& that would probably not be the case for the layout posted by 747).

I hope this helps....

[ringo]

It helps becuase it's logical.
Thanks to 747 and dave W for good posts. This is how it's supposed to be done on F1technical.

Now, I see where no corner springs will obviously allow the roll stiffness of the axle to be based at zero.

This is all good, but we still have the heave spring in combination with the roll bar. This is not independent tuning, the heave is still effectively behaving like a corner spring. And, i think, it doesn't really change the tuning process, it just eliminates a consideration.

What i think this discussion may lead to is which of the K's , by their nature, would best be taken out for a given situation. Because really and truly instead of eliminating the corner spring, it's stiffness can be as small and close to zero as possible, and the heave or arb can be as stiff or soft, to effect the same stiffness as any other system without one of these elements.

Just to prolonging the convo.

F1 will increase by (Kh + Kb)*Z1, whilst F2 will change by (Kh - Kb)*Z1 - similar in principle to the all springs case, since Kc does not affect the load acting on the connected wheel (for my simple test set-up).

total external Reaction on rear axle in pure heave =

F1 + F2 = (Kh + Kb)*Z1 + (Kh - Kb)*Z1 = 2(Kh)Z1

in pure roll=

F1 + F2 = (Kh + Kb)*Z1 + (Kh - Kb)*(-)Z1 = 2(Kb)Z1

Well clearly for the same displacement direction the ARB wont be activated.

The second equation is interesting becuase depending on the orientation of the car; going over a curb for instance would only involve the ARB. This would be softer than the combination of the 2 Kh and Kb or corner spring. However this only happens with equal and opposite displacement.
The heave spring is only eliminated if it's pushed and pulled with the same force by each side; for a zero compression.

For a mix of roll and heave things get complicated, but the roll bar and heave displacements are inextricably linked.

I see where this is going, and it's quite interesting, but it sounds like a preference thing more than a whole new revolution in suspension linking and design.

It will take some serious analyzing to see the pros and the cons.

[DaveW]

I see where this is going, and it's quite interesting, but it sounds like a preference thing more than a whole new revolution in suspension linking and design.

No, not a revolution, but it shifts the boundaries (& it might change the weight & move the c.g. a fraction).

[747heavy]

I´m happy that (thanks to DaveW for his help) we have found some common ground in our conversation.
If such as system, is benifical for you, depends on what you are looking for.
As Scrabs explained, it could be "just" a packaging advantage for some, or a specific characteristic, they would like to achieve for others (such as "zero" roll stiffness, while still maintaining stiffness in heave).
Single wheel bump (or droop) is a combination out of the two "pure" modes roll & heave, so in this case/mode both components (ARB & Heave spring) will contribute force(s) to our system.
Still, if we consider a system with side springs, we are able to seperate our forces in two (pure roll & pure heave) out of three cases, where the side spings would contribute in any (3) cases.
Therefore with a sidespring system, you wont be able to achieve "Zero" roll stiffness.
If you consider this an advantage/improvement will depend from your objectives.

A combined mode, such as single wheel bump or asymetric roll (for example roll with a "zero droop" damper or with rebound springs fitted) will need to be analysed on a case to case basis, as we will have a combination of both components, but this would also be the case, with side springs.

If we consider the analysis of roll and heave spring(stiffness) difficult, then just think about roll & heave damping (velocity sensitive).

Just to round up our conversation, here some photos of other "decoupled" systems.
At the rear they have added a damper to a "monoshock" layout, so that you now can tune roll&heave spring(stiffness) and roll&heave damping individually.

The rear version of the mono suspension is more complicated. It's also a little beefier-looking isn't it. Higher forces back here. The spring and damper are not coilover but are mounted parallel. That makes for quicker changes. The extra damper to the right acts in roll only. The link to the roll damper has a pivot under the other spring. These are Koni 2812 shocks.

(Added 3/2/99) Jeff Braun told me total vertical movement at the rear is typically 0.875 in. max with the mono suspension. Total shuttle movement, side-to-side movement due to roll forces is only 0.075 in.



The photo above shows an alternate Dallara front suspension that uses only one coilover unit. When both the wheels move together in bump or rebound, the pushrods activate the rotating link that moves the spring/damper unit. Roll movement occurs when the pushrods move in different directions. When this happens that link has to slide on its shaft and those dark cylinders on that shaft house springs, probably belleville washers, that add roll stiffness. Under the coilover you can see a cockpit-adjustable blade that adds to the roll stiffness.

photos and text from:
http://insideracingtechnology.com/phnxtst.htm

[747heavy]

Williams gearbox

note the ARB (or lateral mounted torsion bars) and position of the rockers (look for possible space for corner torsion springs)

[ringo]

I´m happy that (thanks to DaveW for his help) we have found some common ground in our conversation.
If such as system, is benifical for you, depends on what you are looking for.
As Scrabs explained, it could be "just" a packaging advantage for some, or a specific characteristic, they would like to achieve for others (such as "zero" roll stiffness, while still maintaining stiffness in heave).
Single wheel bump (or droop) is a combination out of the two "pure" modes roll & heave, so in this case/mode both components (ARB & Heave spring) will contribute force(s) to our system.
Still, if we consider a a system with side springs, we are able to seperate our forces in two (pure roll & pure heave) out of three cases, where the side spings would contribute in any (3) cases.
Therefore with a sidespring system, you wont be able to achieve "Zero" roll stiffness.
If you consider this and advantage/improvement will depend from your objectives.

That's the crux, it depends on what is desired.

Zero roll stiffness, hmmm.... This is only when one wheel is off the ground and ignoring the weight of the un sprung parts? The weight of the lifted side will prevent perfectly zero roll stiffness.
If both wheels are on the ground during roll, any asymmetry in ARB pick up points will impart force on the ARB, so it will be case by case indeed.
The aero loading on top of the car will also quickly return the lifted wheel into the ground, so the moments when zero roll stiffness is achieve are pretty fleeting.

But anyway, the pure modes,as mentioned are where i think removing any spring benefits, as sparingly as it occurs.

Someone needs to interview an F1 suspension engineer. Removal of side springs work of course, as long as the heave spring assumes the role to hold up the car, but most of the curiosity lies with what the engineer is trying to achieve on a whole by removing the corners.

Outside of packaging and tuning for pure modes .. meh.. i like the traditional corner spring setup. There's little difference in mixed modes. Well at least this is how i see it; keep in mind i've never tuned a suspension before, i just know a thing or 2 on vibrations and damping.