PSA: Think twice before lowering your car

[bigjae46]

Interesting perspectives, I never really consider aero for my uses case but it's definitely a thing for serious track guys. At what point did you decide to remove the rear swaybar, and what did you notice as a result? I am considering doing the same but might not have stiff enough springs to enable it.

I removed the rear bar when I built the roll cage but before I put any aero devices on. It adds a lot of rear end grip. A sway bar makes an axle less independent and more of a solid axle...like a Mustang . The sway bar picks up the inside rear wheel when loaded up in a turn so the tire has less grip. No rear sway bar allows the inside wheel to droop.

Some might not like it, might feel lazy or not responsive. You have to be willing to step on the throttle.

I love it! The rear end feels much less skittish. With a ZF LSD I can get on the throttle, get the rear sliding and then it just hooks up and goes.

The flat ride guy has an interesting video on sway bars. I think he mentioned driving with no front bar either on his Miata.

[maw1124]

bigjae46 That’s a great explanation. The sway bar picks up the unloaded wheel… even I can remember that. Looks like the front car has a front sway bar.

Jokes and fun kids, relax.

maw

[Bry5on]

The stock suspension rolls too much which equals more weight transfer so you lose grip as you overwhelm the tires where the weight is transferring to. AND you're transferring weight away from some tires which reduces grip even more. The stock suspension is compromised to add under car clearance (finned diff cover anyone?) and more roll to make the car less twitchy. I can clearly get through a combination of corners (like a switchback) and apply throttle on exit MUCH earlier than a car on a stock suspension..

Just for other readers here who might have read this as factual engineering or physics, roll stiffness is not what determines weight transfer in steady state cornering, cornering load and suspension geometry is the primary determinant of weight transfer, or more specifically, load at each tire in steady state. A firmer set of springs and dampers will transfer load to the outside tires on direction change faster than an equivalently damped softer spring, with the same inputs. It’s the exact opposite of what you’re stating.

That said, fast weight transfer is important for managing quick changes in direction and is definitely a factor worth tuning your setup around.

The stock M3 is mostly compromised because it has to use similar geometry curves and design to the non-M vehicle with which it shared development. As a discrete example, if your front control arms are parallel to the ground at static ride height, a couple things will happen during cornering: 1) your outside tire will lose negative camber as the suspension compresses past a tangent point and 2) your effective spring rate goes down due to the trigonometry and force balance at the tire (this is the roll center and center of mass divergence that folks are referencing, where the cornering tire force has a cross product that directly contributes to compressing the spring, more than just the weight of the car). These are both undesirable effects.

All of this can be designed around, but you’ve really got to know the math and understand what you’re doing. If you don’t have that expertise or experience, you’ll almost certainly end up with a better handling car by managing the roll center instead of the center of mass (read: keep your ride height closer to stock), for any given spring rate. If that’s what you want to optimize around.

It’s been a few years since I’ve had my ground vehicle dynamics books/models out and I designed car suspensions, but the above is pretty basic engineering I feel confident in sharing.

[Bry5on]

I removed the rear bar when I built the roll cage but before I put any aero devices on. It adds a lot of rear end grip. A sway bar makes an axle less independent and more of a solid axle...like a Mustang . The sway bar picks up the inside rear wheel when loaded up in a turn so the tire has less grip. No rear sway bar allows the inside wheel to droop.

Some might not like it, might feel lazy or not responsive. You have to be willing to step on the throttle.

I love it! The rear end feels much less skittish. With a ZF LSD I can get on the throttle, get the rear sliding and then it just hooks up and goes.

The flat ride guy has an interesting video on sway bars. I think he mentioned driving with no front bar either on his Miata.

I used to run a car that had most of its roll stiffness from the springs and I opted for no rear bar at all. If you have enough spring, this is definitely an option. It won’t be the most comfortable ride due to the ride rates, but the math definitely closes and I liked that setup too. It was a little above 2Hz.

[simonnim]

Just for other readers here who might have read this as factual engineering or physics, roll stiffness is not what determines weight transfer in steady state cornering, cornering load and suspension geometry is the primary determinant of weight transfer, or more specifically, load at each tire in steady state. A firmer set of springs and dampers will transfer load to the outside tires on direction change faster than an equivalently damped softer spring, with the same inputs. It’s the exact opposite of what you’re stating.

That said, fast weight transfer is important for managing quick changes in direction and is definitely a factor worth tuning your setup around.

The stock M3 is mostly compromised because it has to use similar geometry curves and design to the non-M vehicle with which it shared development. As a discrete example, if your front control arms are parallel to the ground at static ride height, a couple things will happen during cornering: 1) your outside tire will lose negative camber as the suspension compresses past a tangent point and 2) your effective spring rate goes down due to the trigonometry and force balance at the tire (this is the roll center and center of mass divergence that folks are referencing, where the cornering tire force has a cross product that directly contributes to compressing the spring, more than just the weight of the car). These are both undesirable effects.

All of this can be designed around, but you’ve really got to know the math and understand what you’re doing. If you don’t have that expertise or experience, you’ll almost certainly end up with a better handling car by managing the roll center instead of the center of mass (read: keep your ride height closer to stock), for any given spring rate. If that’s what you want to optimize around.

It’s been a few years since I’ve had my ground vehicle dynamics books/models out and I designed car suspensions, but the above is pretty basic engineering I feel confident in sharing.

What I got out of this is. Keep car stock height but ride on stiffest springs and dampers you can find.

[cobra]

I used to run a car that had most of its roll stiffness from the springs and I opted for no rear bar at all. If you have enough spring, this is definitely an option. It won’t be the most comfortable ride due to the ride rates, but the math definitely closes and I liked that setup too. It was a little above 2Hz.

I took out my rear bar today as a test and I'm amazed how much more grip it has during cornering. I can keep the throttle on so much more. I am just above 1.8Hz.

[Bry5on]

What I got out of this is. Keep car stock height but ride on stiffest springs and dampers you can find.

I prefer ride rates in the 1.4-1.6Hz range for a sporty street car, that’s just my personal preference. Any firmer than that and the car skips around in my local canyons on bumpy sections and gives up traction. My stock cayman and 911 with factory sport suspension options were too stiffly sprung for my canyons as well. I wouldn’t jump to ‘stiffest’ in this case. My homework pointed me to Dinan or Eibach for my preferred ride rates. Since I have a wagon that’s heavier in back, the slightly stiffer rears of the Dinan setup are a great balance for me. There just wasn’t enough front travel so I spaced them up.

[simonnim]

So who did you design suspension for?

[Bry5on]

So who did you design suspension for?

Personal cars and paying friend ‘customers’ - mostly geometry corrections for specific cases. Although a few of my friends that I keep in touch with have gone on to design suspensions for Tesla, Rivian, Nio (the only OEM I actually worked for) professionally

[simonnim]

The stock M3 is mostly compromised because it has to use similar geometry curves and design to the non-M vehicle with which it shared development. As a discrete example, if your front control arms are parallel to the ground at static ride height, a couple things will happen during cornering: 1) your outside tire will lose negative camber as the suspension compresses past a tangent point and 2) your effective spring rate goes down due to the trigonometry and force balance at the tire (this is the roll center and center of mass divergence that folks are referencing, where the cornering tire force has a cross product that directly contributes to compressing the spring, more than just the weight of the car). These are both undesirable effects.

All of this can be designed around, but you’ve really got to know the math and understand what you’re doing. If you don’t have that expertise or experience, you’ll almost certainly end up with a better handling car by managing the roll center instead of the center of mass (read: keep your ride height closer to stock), for any given spring rate. If that’s what you want to optimize around.

.

I had a little more time to think about this. So this goes a bit against big suspension track rats that I've spoken to. By what you said the goal is to prevent losing negative camber, and this can be done either by prevent overall suspension movement or extra camber to calculate in turning weight suspension dip. The first thought in my head is that this can be achieved by progressive springs that most track rats say to avoid and just go straight for linear.

And then my next point is to prevent suspension compression you'd want overall less spring travel, which a shorter spring can achieve and also will lower the car.

Can you explain where these might be wrong? To me it still feels like lowering the car, at least not beyond the point where control arms aren't parallel to the ground is beneficial.

[Bry5on]

I had a little more time to think about this. So this goes a bit against big suspension track rats that I've spoken to. By what you said the goal is to prevent losing negative camber, and this can be done either by prevent overall suspension movement or extra camber to calculate in turning weight suspension dip. The first thought in my head is that this can be achieved by progressive springs that most track rats say to avoid and just go straight for linear.

And then my next point is to prevent suspension compression you'd want overall less spring travel, which a shorter spring can achieve and also will lower the car.

Can you explain where these might be wrong? To me it still feels like lowering the car, at least not beyond the point where control arms aren't parallel to the ground is beneficial.

Your thoughts are on the right track. Extra progressiveness in the springs (or bump stops, as they are also springs) and generally high spring and roll rates combined will be able to combat much of this.

If you’re not driving on a flat road however, you’ll still need suspension travel to allow the car’s wheels to track the road while the car carries its inertia and doesn’t follow the road exactly. The spring is really there to allow the car and the wheel to track different paths (the tire does some of this too). A softer spring rate helps the car keep a more consistent load in the tire contact patch as the wheels move up and down with road imperfections. This ultimately results in more grip on imperfect roads, which all roads are imperfect to some degree. If you have a track only setup, your imperfections will be small, less than 10mm or so, so you can get away with a stiff spring (also, tire choice is very different here). As the spring compresses or unloads with a stiff spring, the force at the tire changes greatly, ultimately not providing a steady load for your tire to resist, which is what you’ll want if you want to be fast.

edit, add: So now we have roads that aren’t perfect, and we must have some soft enough springs, and consequently need suspension travel to allow those springs to do their job. You now have a lower roll rate because of the softer springs, so when you lower your car, even if you start at parallel, your suspension will roll enough to put you into the bad geometry scenario where your control arms are inverting and your roll center starts moving wildly. You basically always end up in a place where you need the ride height if you’re driving on streets that are not perfect. Unless you’re changing control arm and tie rod pickup points…

[Boost]

Interesting info and perspectives in this thread. I am no engineer or race car builder, but if you're selling the idea that a stock ride height car will outperform a lowered car (all else equal), I probably wouldn't buy. But again, I'm just a guy on the internet. Can someone put together a proper real-world comparison between the two? Same driver, same course(s), same conditions, and same car, but with different ride height setups. Not sure how feasible that might be.

Performance aside, I think its fair to say that comfort isn't a big part of the equation (if at all) for most people pursuing performance, and probably shouldn't be. Unless maybe you're LH44/GR63 getting porpoising concussions.. though even in the pinnacle of motorsports, Red Bull and Ferrari have no desire to raise their ride heights.

Personally, hard pass on stock ride height in an E46 M3 off aesthetic alone, but to each his own. I still love all of you, xoxo.

[Casa de Mesa]

I removed the rear bar when I built the roll cage but before I put any aero devices on. It adds a lot of rear end grip. A sway bar makes an axle less independent and more of a solid axle...like a Mustang

I apologize in advance for it because "this isn't M3", but a quick tangent related to the parts I bolded. I'm still sorta on topic. bigjae46 not singling you out on purpose despite me quoting you. Nothing directed at you other than progressing the conversation, my brother in speed.

People like to point to the Mustang as the poster child for bad handling (myself included), in large part because of its Model T suspension design and propensity for snap oversteer. That reputation is due to its stock 4-arm suspension design which binds under heavy lateral/cornering loads, effectively taking the spring rate way up. However, there are a lot of suspension designs built around a live axle rear that are pretty amazing. Like a Watts link or panhard bar combined with a torque arm.

The beauty of a properly setup solid/live axle is that you don't need to worry about the camber changing - BOTH rear tires stay planted on the ground regardless of what the chassis is doing above it. This equates to lots of grip, better braking, a more predictable behavior at the limit (oh the irony) and the ability to put the power down incredibly well. Especially with a torque arm which loads those rear tires up like John Pinette's plate at a Chinese buffet.

Here are 3 pics, all in the same relatively high-G corner, that illustrate two points:

1) Removing the sway bar on the M3 allows the inside tire to stay in contact with the road. With the solid axle, it just happens naturally. Despite its reputation, the stick in the rear can handle like a mofo.

2) The downsides to static negative camber on a McPherson strut design.

First pic in mid-corner, look how flat the rear axle is compared to the inside front tire. Same thing with the second pic on corner exit.



Third pic is the same corner but shot from the front. It illustrates my 2nd point. On this car, this is pretty perfect [front] negative camber for that track at that corner - even still, the inside tire is marginalized (pic 1 and 2) due to that static negative camber. But look at the contact patch on the rear inside tire...

I took out my rear bar today as a test and I'm amazed how much more grip it has during cornering. I can keep the throttle on so much more. I am just above 1.8Hz.

I love that you did this!! How do you know you're at 1.8Hz?

[simonnim]

I apologize in advance for it because "this isn't M3", but a quick tangent related to the parts I bolded. I'm still sorta on topic. bigjae46 not singling you out on purpose despite me quoting you. Nothing directed at you other than progressing the conversation, my brother in speed.

People like to point to the Mustang as the poster child for bad handling (myself included), in large part because of its Model T suspension design and propensity for snap oversteer. That reputation is due to its stock 4-arm suspension design which binds under heavy lateral/cornering loads, effectively taking the spring rate way up. However, there are a lot of suspension designs built around a live axle rear that are pretty amazing. Like a Watts link or panhard bar combined with a torque arm.

The beauty of a properly setup solid/live axle is that you don't need to worry about the camber changing - BOTH rear tires stay planted on the ground regardless of what the chassis is doing above it. This equates to lots of grip, better braking, a more predictable behavior at the limit (oh the irony) and the ability to put the power down incredibly well. Especially with a torque arm which loads those rear tires up like John Pinette's plate at a Chinese buffet.

Here are 3 pics, all in the same relatively high-G corner, that illustrate two points:

1) Removing the sway bar on the M3 allows the inside tire to stay in contact with the road. With the solid axle, it just happens naturally. Despite its reputation, the stick in the rear can handle like a mofo.

2) The downsides to static negative camber on a McPherson strut design.

First pic in mid-corner, look how flat the rear axle is compared to the inside front tire. Same thing with the second pic on corner exit.



Third pic is the same corner but shot from the front. It illustrates my 2nd point. On this car, this is pretty perfect [front] negative camber for that track at that corner - even still, the inside tire is marginalized (pic 1 and 2) due to that static negative camber. But look at the contact patch on the rear inside tire...




I love that you did this!! How do you know you're at 1.8Hz?

This is awesome illustration of how negative camber is affecting at turn in.

[simonnim]

Interesting info and perspectives in this thread. I am no engineer or race car builder, but if you're selling the idea that a stock ride height car will outperform a lowered car (all else equal), I probably wouldn't buy. But again, I'm just a guy on the internet. Can someone put together a proper real-world comparison between the two? Same driver, same course(s), same conditions, and same car, but with different ride height setups. Not sure how feasible that might be.

Performance aside, I think its fair to say that comfort isn't a big part of the equation (if at all) for most people pursuing performance, and probably shouldn't be. Unless maybe you're LH44/GR63 getting porpoising concussions.. though even in the pinnacle of motorsports, Red Bull and Ferrari have no desire to raise their ride heights.

Personally, hard pass on stock ride height in an E46 M3 off aesthetic alone, but to each his own. I still love all of you, xoxo.

I think Bry5on implies that because of imperfect streets (pot holes, cracks, light up road dividers) that more travel for street driving is recommended vs the track where he said we can get away with 10mm travel and thus like lowering the car to prevent overall travel to retain correct negative camber during turn in with stiffer springs will be better.

Correct me if I misunderstood.