Sway bar vs Springs to Control Body Roll

I know everything is a compromise, and it seems to me you control body roll with a combination of spring rate and sway bar. Is there an advantage or disadvantage to using stiffer springs vs a stiffer sway bar to address this? Of course stiffer springs on deteriorated roads is going to be more punishing, but leaving that aside, which is "better?" For example, would it be better to go with a 350 lb/in front spring and a stiffer sway bar, or a 600 lb/in spring and less sway bar? I ask this with my Miata in mind

We all know the good people at Mazda went with soft springs and bump stops as integral parts of the suspension, so we don't necessarily need to get into that.

The advantage is pitch and dive control and weight reduction.

The disadvantage is ride quality.

IMO, go stiff enough on the springs to get responsive enough front / back weight transfer and to avoid needing a huge bar on the drive end. If your shocks are well matched to the springs, it should still ride acceptably.

Springs are "better."

Springs affect most aspects of the car's suspension. Sway bars isolate roll stiffness alone. So if you only want to decrease body roll without any other changes, go with roll bars. If you want to decrease pitch and dive, or change the ride height, or modify how the car deals with single or multi-wheel bumps, take the springs into consideration.

Keep in mind that springs have to be appropriately damped, so a change in spring rate may also require a chance in shocks. Also keep in mind that it's very easy to adjust the front/rear roll stiffness with sways, which means the handling balance.

Also, we're talking about a Miata here. It's a pretty well-developed platform. You may do just as well by saying "here's what I want to do, what do you suggest"?

I'm a big fan of barless setups with light cars,I bought my NA with upgraded sways front and rear and a 500lb front 350lb rear spring setup.

Autox only and went straight to an 800/500 spring setup,figured I'd leave the bars alone and see how it worked figuring the bars would be too big for those rates.

Nope,ordered up the bigazz front bar when I got back from the event.

Miatas are pretty rolly & undersprung by nature (NA/NB anyway). As stated above if you're looking to tame roll & tighten turn-in only, go for the bars. But don't be surprised if your old Showas get confused on washboard stuff, the bars add to the spring rate on uneven L/R surfaces. So you're in for new shocks as well if you have choppy roads.

Otherwise, suck it up & just go for the matched set. Springs/shocks only won't tame the roll.

Keith explaned it well, Springs support the weight Sway bars control roll But By nature also affect spring action, then the shocks control Jounce and rebound On our short track cars we hooked the sway bar with a link on the left side and allowed it to be free on the right, when in compression it touched the lower right control arm and then became actuated. but they still are Two different items and do two different things.

I'm in the soft spring bigger bar camp.

As an aside when the suspension on our D-sports racer was redone, the car had no sway bars. The tube frame chassis was no more than 20" wide the car was 72" wide. Basically the design followed classic Carroll Smith that the longer suspension arms the better. Note the car set the class lap record locally. Of all the single seat cars I've driven it was one of the very few that had no issues pounding over the kerbs.

Now the issue here is with a street car like a Miata it's simply not practical.

Tom

I already have a nice set of revalved Bilstein/coilovers that have adequate damping for the high rate springs on the car. Not too much roll even on the stock bars now. I am pretty happy with it as it sits, DD/autocross at least a half dozen weekends a year.

I ask more out of curiosity than looking for specific direction. Lot of accumulated knowledge up in here.

A couple of things nobody has mentioned about bars that you might be interested in.

1. A swaybar works by converting body roll into weight transfer. Since the tire's coefficient of friction decreases with load, this means the mechanical grip of the entire system decreases with bars.

2. The bigger the bars, the more you've effectively converted your suspension into a very heavy and complicated solud axle system. No bars at all equal completely independent suspension. Infinitely stiff bars are, for practical purposes, stick axles.

In reply to DaewooOfDeath:

Good points. Another is that since a sway bar is just another spring, such that your spring rate is now different in ride (and pitch) than roll, that means your damping can be 'ideally' tuned for one, the other, or neither...But not both.

I'm not much a fan of my RX-7s suspension anymore. I removed the sway bars to make servicing easier, but the springs/damping required add too many compromises.

If you're not having pitch problems, sway bars and softer springs are better IMO. Softer springs allow softer dampers which means the tires follow the road better. Smooth inputs by the driver mean nothing if any little bump or dip causes massive load variations because the dampers are really stiff.

DaewooOfDeath wrote: A couple of things nobody has mentioned about bars that you might be interested in. 1. A swaybar works by converting body roll into weight transfer.

This is a false assumption. Weight transfer happens strictly as a function of center of gravity vs. track width. Different roll stiffness front to rear alters the roll stiffness distribution, and both springs and sway bars affect this. If you stiffen the springs to make up for a lack of sway bars, it's zero-sum.

2. The bigger the bars, the more you've effectively converted your suspension into a very heavy and complicated solud axle system. No bars at all equal completely independent suspension. Infinitely stiff bars are, for practical purposes, stick axles.

You don't understand how solid axles work They are the opposite of infinitely stiff. Solid axles have the downside of higher unsprung weight, but they articulate MORE than independent suspension does. This is because the springs and shocks are inboard of the contact patches, so there is a lever effect against the other side.

Knurled wrote:

DaewooOfDeath wrote: A couple of things nobody has mentioned about bars that you might be interested in. 1. A swaybar works by converting body roll into weight transfer.

This is a false assumption. Weight transfer happens strictly as a function of center of gravity vs. track width. Different roll stiffness front to rear alters the roll stiffness *distribution*, and both springs and sway bars affect this. If you stiffen the springs to make up for a lack of sway bars, it's zero-sum.

Exactly. It's a very common error to make. A lot of it comes out of the fact that you can use sway bars to affect the distribution of weight transfer. Put a stiffer sway bar at one end, that end gets a greater proportion of the weight transfer and it drops in grip. But the other end gets less weight transfer and gains grip. The total transfer stays the same. If you stiffen both ends by the same proportion, the weight transfer distribution stays the same and the grip level stays the same. Also, this happens with springs as well, it doesn't matter where your roll stiffness comes from.

This is ignoring edge cases (running out of travel) or effects from suspension geometry changes. Decreasing body roll will have an effect there.

I don't think I've ever dealt with a set of sway bars so stiff that it caused problems with damping.

Keith Tanner wrote:

Knurled wrote:

DaewooOfDeath wrote: A couple of things nobody has mentioned about bars that you might be interested in. 1. A swaybar works by converting body roll into weight transfer.

This is a false assumption. Weight transfer happens strictly as a function of center of gravity vs. track width. Different roll stiffness front to rear alters the roll stiffness *distribution*, and both springs and sway bars affect this. If you stiffen the springs to make up for a lack of sway bars, it's zero-sum.

Exactly. It's a very common error to make. A lot of it comes out of the fact that you can use sway bars to affect the distribution of weight transfer. Put a stiffer sway bar at one end, that end gets a greater proportion of the weight transfer and it drops in grip. But the other end gets less weight transfer and gains grip. The total transfer stays the same. If you stiffen both ends by the same proportion, the weight transfer distribution stays the same and the grip level stays the same. Also, this happens with springs as well, it doesn't matter where your roll stiffness comes from. This is ignoring edge cases (running out of travel) or effects from suspension geometry changes. Decreasing body roll will have an effect there. I don't think I've ever dealt with a set of sway bars so stiff that it caused problems with damping.

No, guys, I understand about front to rear balance. It doesn't matter if we're putting swaybars on a Segwey, it's still going to turn roll into weight transfer, at least as far as I understand. (I'm plagarizing all this from Steve Hoelscher, btw)

Think of it this way, you are being pushed to one side like the black dude #44 in this picture.

In order to resist the rolling force of the white dude, you effectively do what a swaybar in a car does - namely, you increase the amount of force on your outside leg/wheel, decrease the amount of force on your inside leg/wheel and fight against the rolling force, whether cornering or a lineman.

The same thing happens with a swaybar. The links on the outside tire's side push down on the strut/control arm/whatever and the links on the inside tire's side lift up. In other words, they convert roll into weight transfer.

So, since coefficient of friction decreases with load, you always want your side to side weight split as even as possible. The swaybars, by converting roll to weight transfer, make the side to side weight split more severe and thus decrease the total amount of grip in the system.

Also, what I meant about infinitely stiff swaybars is that any impact on one side of the car is going to be transferred to the other side. With no swaybars, you have total isolation side to side, which is nice for bumpy roads/tracks/whatever. Though, to be fair, I should have said that infinitely stiff bars basically make your rear suspension the same as a golf cart.

Keith Tanner wrote: I don't think I've ever dealt with a set of sway bars so stiff that it caused problems with damping.

Hoo boy. This comment reminded me of that time that Golf of mine elected to liberate one of its rear shocks' oil. In that timeframe, VW had a necessarily hyper-stiff rear "sway bar" (twist beam suspension) and a tall, lightweight chassis with low roll centers - therefore, it would lift a rear tire at slightest provocation. To combat any oversteer effects, along with the effects of the beam bending under load, VW employed an interesting suspension bushing with deliberate side play and an internal ramp, to increase outside wheel toe-in on cornering. And they use a TON of static toe-in and negative camber. Lots. Ridiculous tire-wear amounts of toe-in and camber. The reason for the stupid "hellaflush" fad amounts of toe-in and camber.

So, to recap: Lots of static toe-in on both sides. Very stiff spring connecting the two sides. Remove damping from one side.

Every time the right rear hit a bump, it would jackhammer. Every time it did that, the large amount of toe-in on the left side would yaw the car to the left until the right rear settled down.

That was a very entertaining three days of life before my new shocks came in. I installed them outside. At night. In the rain.

DaewooOfDeath wrote: Think of it this way, you are being pushed to one side like the black dude #44 in this picture. In order to resist the rolling force of the white dude, you effectively do what a swaybar in a car does - namely, you increase the amount of force on your outside leg/wheel, decrease the amount of force on your inside leg/wheel and fight against the rolling force, whether cornering or a lineman. The same thing happens with a swaybar. The links on the outside tire's side push down on the strut/control arm/whatever and the links on the inside tire's side lift up. In other words, they convert roll into weight transfer.

And it's the same thing if you are using springs for roll stiffness. Doesn't matter. The weight transfer is because of CG vs. track width, no matter if the car is rolling 3 degrees or 15 degrees. The weight transfer is because of physics, the suspension is only there for motion control.

So, since coefficient of friction decreases with load,

This is also not quite correct. The relation between load and grip is nonlinear. The more the difference between the two tires is, the less the TOTAL grip is. Either way, this is something that is defined by the center of gravity versus track width. No matter how you tune the suspension, you aren't going to make a 911 handle like a Civic.

It can be the opposite on loose surfaces, too. You may want to increase weight transfer in order to get more grip.

I'm pretty sure it's not the same, but let's try out your version.

If we have a 100% spring set up, is there any mechanical device decreasing the effective at-the-wheel-rate on the inside wheel and increasing the effective rate on the outside wheel during a cornering event?

Spring rate has nothing to do with weight transfer except for how it is distributed front to rear.

This, I think, is your hangup. You're equating roll resistance with weight transfer. They are unrelated. Roll resistance is only useful for body motion control.

In reply to Knurled:

Roll resistance and weight transfer have (almost) no relationship. Unless you resist roll with swaybars, in which case the swaybar converts roll to weight transfer.

Here's what I'm talking about.

http://www.mr2oc.com/showthread.php?t=298272

Roll stiffness does affect weight transfer - the total weight transfer is split across the front and rear axles, and distributed according to the relative front and rear roll stiffness. So you can't change the total weight transfer with roll stiffness, but you can decide which end gets it.

Body roll is a visible manifestation of weight transfer, not a cause.

Physics, graphs of grip vs load, etc: Phyisics of handling chapter from some book

No springs, soft springs, super stiff springs, no bars or sway bars the size of sky scrapers I beams, Mr Newton says total side to side weight transfer won't change.

Me newton is a pretty smart dude.

DaewooOfDeath wrote: In reply to Knurled: Roll resistance and weight transfer have (almost) no relationship. Unless you resist roll with swaybars, in which case the swaybar converts roll to weight transfer.

Negative. The sway bars resist body roll. The weight transfer doesn't change. Weight tranfer is strictly a matter of center of gravity versus track width. Increasing cornering forces, increasing CG height, or narrowing track width will increase weight transfer, nothing else.

There are very specific effects that you can MAKE with sway bars regarding the front to rear roll stiffness distribution, that are easier to do with sway bars than springs, but it doesn't affect weight tranfer the way I think you think it does.

Keith Tanner wrote: Body roll is a visible manifestation of weight transfer, not a cause.

I understand this and it's not what I'm talking about.

(I really do recommend you read Steve Hoelscher's set up thing - http://www.mr2oc.com/showthread.php?t=298272.)

Knurled wrote: The sway bars resist body roll.

I'd ask Keith and Knurled to forget about body roll for a moment and just imagine I'm standing in front of your Miata.

Pretend I have really long arms and I'm able to grab the lower control arms on both the right and left sides of your car at the same time.

If I push down on the left control arm with 100 lbs of force, will that force be transferred to the left tire?

If I pull up on the right control arm at the same moment with 100 lbs of force, will that force be transferred away from the right tire?